diff --git "a/FNE4T4oBgHgl3EQffg3m/content/tmp_files/load_file.txt" "b/FNE4T4oBgHgl3EQffg3m/content/tmp_files/load_file.txt" new file mode 100644--- /dev/null +++ "b/FNE4T4oBgHgl3EQffg3m/content/tmp_files/load_file.txt" @@ -0,0 +1,814 @@ +filepath=/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf,len=813 +page_content='Counterfactual Explanations for Concepts in ELH Leonie Nora Sieger Paderborn University Paderborn, Germany Stefan Heindorf Paderborn University Paderborn, Germany Lukas Blübaum Paderborn University Paderborn, Germany Axel-Cyrille Ngonga Ngomo Paderborn University Paderborn, Germany ABSTRACT Knowledge bases are widely used for information management on the web, enabling high-impact applications such as web search, question answering, and natural language processing.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' They also serve as the backbone for automatic decision systems, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' for medi- cal diagnostics and credit scoring.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' As stakeholders affected by these decisions would like to understand their situation and verify fair decisions, a number of explanation approaches have been proposed using concepts in description logics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' However, the learned concepts can become long and difficult to fathom for non-experts, even when verbalized.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Moreover, long concepts do not immediately provide a clear path of action to change one’s situation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Counterfactuals answering the question “How must feature values be changed to ob- tain a different classification?”' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' have been proposed as short, human- friendly explanations for tabular data.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In this paper, we transfer the notion of counterfactuals to description logics and propose the first algorithm for generating counterfactual explanations in the description logic ELH.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Counterfactual candidates are generated from concepts and the candidates with fewest feature changes are selected as counterfactuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In case of multiple counterfactuals, we rank them according to the likeliness of their feature combinations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For evaluation, we conduct a user survey to investigate which of the generated counterfactual candidates are preferred for explanation by participants.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In a second study, we explore possible use cases for counterfactual explanations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' CCS CONCEPTS Information systems → Semantic web description languages;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' • Computing methodologies → Description logics;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' • Human- centered computing → Empirical studies in HCI.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' KEYWORDS Description Logic, Knowledge Graphs, Machine Reasoning, XAI, Semantic Web 1 INTRODUCTION Knowledge bases are commonly used in web applications, includ- ing information retrieval [30], information generation [27], web search [5] and question answering [12].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' A great share of work is concerned with securing correctness and completeness of knowl- edge bases [9, 20, 28] which often involve machine learning-based classification.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Concepts in description logics (DLs) can serve as transparent, white-box models for binary classification.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Many ap- proaches to learn concepts from positive and negative examples have been proposed [7, 11, 14, 17, 18, 31].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Such DL concepts can Person 1 Female Person 2 hasChild Male Person 3 married Female Person 1 Female Person 2 Male Person 3 married Female Person 1 Female Person 2 Male hasChild Person 3 married Female Figure 1: The concept Female⊓∃hasChild.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='⊤ classifies Person 1 as a mother in KB 1 (green), while its corresponding counterfactuals in KBs 2 and 3 (red) are not classified as such.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 1 Figure 1: The concept Female ⊓ ∃hasChild.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='⊤ classifies Per- son 1 as a mother in KB 1 (green), while its corresponding counterfactuals in KBs 2 and 3 (red) are not classified as such.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' directly be mapped to expressions in the Web Ontology Language (OWL) used in the Semantic Web [16, 24].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' This makes DL concept learning a convenient machine learning tool to apply on webbased knowledge graphs like DBpedia [3], Wikidata [35], or YAGO [34], or semantic website data, as proposed by schema.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org [15].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Explaining algorithmic decisions of machine learning models to stakeholders has become increasingly important [1, 2, 13, 25]: subjects affected by model decisions would like to understand their situation and verify fair decisions;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' data scientists would like to debug and improve the model;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' regulatory entities would like to check the compliance with laws and regulations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' However, as concepts increase in length and complexity, their practical utility decreases, making it increasingly difficult for stake- holders to understand, contest or alter decisions [36].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' To increase acceptance and trust of stakeholders, counterfactual explanations have been proposed as a form of short, actionable explanations [26].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Counterfactual explanations focus on an antecedent that would have caused a different outcome (classification) had it been the case [32].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Given a set of features 𝐴 and a classification 𝐵, a counter- factual statement takes on the form “If 𝐴 had not been true, then the classification would not have been 𝐵”, where in the current classification to be explained, both 𝐴 and 𝐵 are true.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In this paper, we transfer the notion of counterfactuals to DLs and propose the first algorithm for generating counterfactual explana- tions in the DL ELH.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' ELH as it is an important description logic for many applications including the medical ontology Snomed [6] arXiv:2301.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='05109v1 [cs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='AI] 12 Jan 2023 Sieger et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Table 1: ELH description logic constructs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For further de- tails, we refer to Lehmann and Turhan [23].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Syntax Semantics Construct ⊤ ΔI top concept 𝐶, 𝐷 𝐶I, 𝐷 I ⊆ ΔI atomic concepts 𝑟,𝑠 𝑟 I,𝑠I ⊆ ΔI × ΔI atomic roles 𝐶 ⊓ 𝐷 𝐶I ∩ 𝐷 I intersection ∃𝑟.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='𝐶 {𝑥 ∈ ΔI|∃𝑦 ∈ ΔI existential restriction Concepts with (𝑥,𝑦) ∈ 𝑟 I ∧ 𝑦 ∈ 𝐶I} 𝐶(𝑥) 𝑥 I ∈ 𝐶I concept assertion ABox 𝑟 (𝑥,𝑦) (𝑥 I,𝑦I) ∈ 𝑟 I role assertion 𝐶 ⊑ 𝐷 𝐶I ⊆ 𝐷 I concept subsumption TBox 𝑟 ⊑ 𝑠 𝑟𝐼 ⊆ 𝑠I role subsumption and protein-protein interaction networks [19].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Given an individual 𝑥 with concise bounded description CBD(𝑥) and a concept 𝐶 that holds for 𝑥, we generate counterfactual candidates 𝑥 ′ with concise bounded descriptions CBD(𝑥 ′) for which the concept does not hold.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Figure 1 shows an example.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In line with Wachter et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [36], we define counterfactuals as those candidates that are most similar to the original individual.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Next, we rank them according to the plausibility that they appear in the real world, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=', their “likeliness of the combination of their features.” We conduct a user survey to investigate if the selected counterfactuals are indeed preferred by users.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Finally, we conduct a study to explore possible uses of counterfactual explanations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In what follows, Section 2 introduces preliminares, Sec- tion 3 describes related work, Section 4 introduces our al- gorithm to generate counterfactuals, and Sections 5 and 6 conduct user studies to investigate user preferences and use cases.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Finally, Section 7 discusses limitations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' All data, code and materials needed for reproducing this work can be found at https://anonymous.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='4open.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='science/r/Counterfactual- Explanations-ELH-EBDA/.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2 PRELIMINARIES We give a brief overview of the description logic ELH, knowledge bases, their triple representation, the concise bounded description of entities, and the closed-world assumption.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For further details, we refer the reader to Baader et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [4], Brandt [6], Lehmann and Turhan [23].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 The Description Logic ELH.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In DLs [4], knowledge is repre- sented by concept descriptions built from atomic concepts 𝐶, 𝐷 ∈ 𝑁𝐶 and roles 𝑟,𝑠 ∈ 𝑁𝑅, where 𝑁𝐶 and 𝑁𝑅 are finite sets of concept and role names.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' As shown in Table 1, a concept in the description logic ELH [6, 23] can consist of the the top-concept (⊤), inter- sections (𝐶 ⊓ 𝐷), and existential restrictions (∃𝑟.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='𝐶).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Their seman- tics is defined in terms of the interpretation I = (ΔI, ·I) which consists of the non-empty set ΔI, called interpretation domain, and the function ·I, called interpretation function, that assigns each 𝐴 ∈ 𝑁𝐶 a set 𝐴I ⊆ ΔI and each 𝑟 ∈ 𝑁𝑅 a binary relation 𝑟 I ⊆ ΔI × I [22, 23].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Following Lehmann and Hitzler [22], we extend the definition to also assigns each individual name 𝑥 ∈ 𝑁𝐼 an element 𝑥 I ∈ ΔI.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2 Knowledge Base.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' A Knowledge Base (KB) K consists of a TBox T and ABox A where the TBox defines concepts by means of concept inclusion (𝐶 ⊑ 𝐷) and role inclusion axioms (𝑟 ⊑ 𝑠) and the ABox defines individuals 𝑥 ∈ 𝑁𝐼 and their relations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' An interpretation I is a model of the knowledge base K iff it satisfies all axioms in the TBox and ABox.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' An individual 𝑥 ∈ 𝑁𝐼 is an instance of a concept 𝐶 with respect to K, written K |= 𝐶(𝑥) iff in all models I of K, we have that 𝑎I ∈ 𝐶I.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We say 𝐶 holds for 𝑥 in K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Otherwise, we write K ̸|= 𝐶(𝑥) and say that 𝑥 does not hold for C.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 Triple Representation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The ABox can be represented in the form of subject-predicate-object triples as follows: For each individ- ual 𝑥 ∈ 𝑁𝐼 , we obtain all atomic concepts 𝐶 ∈ 𝑁𝐶 and all relations 𝑟 ∈ 𝑁𝑅 such that K |= 𝐶(𝑥) or K |= 𝑟 (𝑥,𝑥 ′) for an individual 𝑥 ′ holds, respectively.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Then each role assertion 𝑟 (𝑥,𝑥 ′) corresponds to a triple (𝑥,𝑟,𝑥 ′) and each concept assertion 𝐶(𝑥) to a triple (𝑥, rdf:type,𝐶).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='4 Concise Bounded Description.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We define the concise bounded description2 CBD(𝑥) of an individual 𝑥 as the set of all triples that contain 𝑥 as a subject (both of the form (𝑥,𝑟,𝑥 ′) and (𝑥, rdf:type,𝐶)).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='5 Closed-world assumption.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' As is common practise for con- cept learning, we employ closed-world semantics [7, 14, 22]: For each individual 𝑥 ∈ 𝑁𝐼 and concept 𝐶 ∈ 𝑁𝐶, either K |= 𝐶(𝑥) or K ̸|= 𝐶(𝑥) holds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In the latter case, we say that 𝐶(𝑥) does not hold.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' This approach allows to handle ontologies as a database and is com- patible with the state-of-the-art concept learners DL-Learner [21] and EvoLearner [14], whose classifications we would like to explain with our approach.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 3 RELATED WORK In the following, we introduce related work on counterfactuals and similarity measures for description logics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Counterfactuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Stepin et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [32] survey research papers on contrastive and counterfactual explanation generation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' They com- pare definitions, summarize methods, and discuss how methods are grounded in theoretical approaches.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' While some approaches are based on white-box models such as decision trees [33], none of them generates counterfactuals for individuals in description logics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' According to Wachter et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [36], counterfactual explanations are statements of the form “Score 𝑝 was returned because 𝑉 had values (𝑣1, 𝑣2, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=') associated with them.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' If 𝑉 had values (𝑣′ 1, 𝑣′ 2, .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=') and all other variables had remained constant, score 𝑝′ would have been returned.” While many such explanations are possible, an ideal counterfactual alters values as little as possible and represents the ‘closest world’ under which score 𝑝′ is returned instead of 𝑝.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 1Note that the |= operator expresses reasoning and we employ a reimplementation of the fast instance checker [22] which follows the closed-world assumption.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For example, the reasoner, derives 𝐷 (𝑥) from the assertions 𝐶 (𝑥) and 𝐶 ⊑ 𝐷.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2Originally, the concise bounded description (CBD) was defined for RDF, cf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=', https: //www.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='w3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/Submission/CBD/ Counterfactual Explanations for Concepts in ELH Dandl et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [10] generalize this idea and take further criteria into account.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Given a statement of the form “If 𝑋 had not occurred, 𝑌 would not have occurred”, they solve a multi-objective optimization problem with four objectives: (1) the prediction 𝑦 of the counter- factual 𝑥 ′ should be as close as possible to the desired prediction 𝑦′;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' (2) The counterfactual 𝑥 ′ should be as similar as possible to the instance 𝑥;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' (3) feature changes should be sparse;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' (4) the counter- factual should have likely feature values/combinations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We use the approach by Dandl et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [10] as a basis for our approach and adapt it to description logics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Similarity Measures for Description Logics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' At the core of coun- terfactuals is the notion of ‘closest world’ and the question of how to operationalize this notion for description logics arises.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Lehmann and Turhan [23] create a framework for semantic-based similarity measures for ELH-concepts.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' They define desirable properties and present a framework to construct similarities measures fulfilling many of the properties.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' However, we need to define a similarity measure for individuals and not for concepts.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 4 COUNTERFACTUALS IN ELH Following Wachter et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [36] and Dandl et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [10] who defined counterfactuals for black-box machine learning models with fixed- size input vectors, we transfer their definition to individuals in description logic.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Given an individual 𝑥 ∈ 𝑁𝐼 and an (atomic or non-atomic) concept 𝐶, 𝑥 ′ is a counterfactual candidate of 𝑥 with respect to 𝐶 iff K |= 𝐶(𝑥) and K ̸|= 𝐶(𝑥 ′) or K ̸|= 𝐶(𝑥) and K |= 𝐶(𝑥 ′) (1) i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=', the predictions of 𝑥 and 𝑥 ′ differ with respect to 𝐶.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Without loss of generality, we focus on the case that 𝐶 holds in the original KB and should not hold for the counterfactual.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The individual 𝑥 ′ does not necessarily need to exist in the original KB K and can be a new hypothetical individual in an alternative knowledge base K′ defined over the same atomic concepts and roles as K.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Let 𝛿(𝑋,𝑌) := |𝑋 \\𝑌 | + |𝑌 \\ 𝑋 | be the edit distance between two arbitrary sets counting the number of additions and removals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' A counterfactual candidate 𝑥 ′ is called a counterfactual of 𝑥 iff the edit distance 𝛿(CBD(𝑥), CBD(𝑥 ′)) (2) between the concise bounded descriptions of the individuals 𝑥 and 𝑥 ′ is minimal among all counterfactual candidates.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' If the concept is ⊤, no counterfactual exists in ELH and we define the edit distance as infinite.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Among multiple counterfactuals, we prefer counterfactuals which are likely and we experimented with two variants of mea- suring likeliness: Let 𝑁𝐼 be the set of all existing negative individ- uals 𝑥 ∈ 𝑁𝐼 with K ̸|= 𝐶(𝑥).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Moreover, given an individual 𝑥, let 𝐶𝑅(𝑥) := {𝐶 ∈ 𝑁𝐶 |K |= 𝐶(𝑥)} ∪ {𝑟 ∈ 𝑁𝑅|∃𝑧 ∈ 𝑁𝐼 : K |= 𝑟 (𝑥,𝑧)} be the set of all atomic concepts and roles of 𝑥.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We define the min- likeliness, variantmin of a counterfactual 𝑥 ′ with K ̸|= 𝐶(𝑥 ′) as the minimal distance: min 𝑥 ∈𝑁𝐼 𝛿(CR(x′), CR(x)).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' (3) The rationale is that a counterfactual is likely if another negative individual in the KB exists that is similar to it.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' As an alternative, variantmean, we define the mean-likeliness of a counterfactual 𝑥 ′ with K ̸|= 𝐶(𝑥 ′) as the average distance: 1 |𝑁𝐼 | ∑︁ 𝑥 ∈𝑁𝐼 𝛿(CR(x′), CR(x)) (4) The rationale is that a counterfactual is likely if it, on average, is similar to the other negative individuals in the KB.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We decided not to take the objects of role assertions into account (as would be done by the CBD) because the objects of roles can be rather distinct in real-world knowledge bases, so that individuals rarely share both roles and role objects.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For example, in the family ontology (cf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=', Figure 1), it is unlikely that two individuals are married to the same person.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In practice, the best choice of similarity function depends on the dataset and use case.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 Comparison to Dandl et al.’s Counterfactual Objectives Our definitions contain the four objectives 𝑜1–𝑜4 by Dandl.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' (𝑜1: Prediction) “A counterfactual instance produces the prede- fined prediction as closely as possible.” Our predefined classification of a counterfactual 𝑥 ′ is the opposite of the classification of 𝑥 as pointed out in Equation 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' (𝑜2, 𝑜3: Similarity, Sparseness) “A counterfactual should be as similar as possible to the instance regarding feature values” and “change as few features as possible.” Since all our CBDs (or “fea- tures”) are discrete and do not contain numeric values, 𝑜2 and 𝑜3 collapse.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We define the similarity according to the edit distance in Equation 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' (𝑜4: Likeliness) “Counterfactuals should have likely feature val- ues/combinations”.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We compute the likeliness/plausibility of feature values as being close to existing individuals in the knowledge base, see Equations 3 and 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2 Generation of Counterfactuals Let K be a knowledge base, 𝐶 an ELH concept, and 𝑥 an indi- vidual 𝑥 for which the concept holds, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=', K |= 𝐶(𝑥).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Let 𝐷 ∈ 𝑁𝐶 be a named concept and 𝑟,𝑟 ′ ∈ 𝑁𝑅 be role names.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We construct counterfactuals 𝑥 ′, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' individuals most similar to 𝑥 for which𝐶(𝑥 ′) cannot be inferred.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For this, we assume K’s TBox to contain only subsumptions 𝐶 ⊑ 𝐷 and 𝑟 ⊑ 𝑟 ′ of atomic concepts 𝐶, 𝐷 ∈ 𝑁𝐶 and roles 𝑟,𝑟 ′ ∈ 𝑁𝑅.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' This assumption holds for a vast array of real-world knowledge bases [37].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Algorithms 1 and 2 describe our approach to generate counter- factuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' First, a reasoner infers as many concept and role assertions as possible taking subsumptions into account (Algorithm 1, line 5).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Next, we create a deep copy K𝑖 of K for each subconcept 𝐶𝑖 of 𝐶 = 𝐶1 ⊓ 𝐶2 ⊓ .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' ⊓ 𝐶𝑛 (line 7).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The method get_individual retrieves the individual 𝑥𝑖 from K𝑖 corresponding to 𝑥 in K (line 8).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We turn the 𝑥𝑖 into a counterfactual candidate with K𝑖 ̸|= 𝐶(𝑥 ′) (line 9) by removing axioms such that the subconcept 𝐶𝑖 does not hold anymore (and hence the concept 𝐶 does not hold for 𝑥𝑖 any- more).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Algorithm 2 describes the generation of a single counter- factual candidate.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Having generated the candidates, the candidates Sieger et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 1 Input: KB K, Concept 𝐶 = 𝐶1 ⊓ 𝐶2 ⊓ .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' ⊓ 𝐶𝑛 with 𝑛 ≥ 1, Individual 𝑥 such that K |= 𝐶(𝑥) 2 Output: Counterfactuals of individual 𝑥 w.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='r.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='t.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 𝐶 3 Function gen_counterfactuals(K, 𝐶, 𝑥): 4 𝑐𝑎𝑛𝑑𝑖𝑑𝑎𝑡𝑒𝑠 ← [ ] 5 K ← reasoner(K) 6 for each 𝐶𝑖 do 7 K𝑖 ← deepcopy(K) // Get copy of 𝑥 in K𝑖 8 𝑥𝑖 ← get_individual(K𝑖,𝑥) 9 K𝑖,𝑥𝑖 ← gen_candidate(K𝑖, 𝐶𝑖, 𝑥𝑖) 10 𝑐𝑎𝑛𝑑𝑖𝑑𝑎𝑡𝑒𝑠.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='append((K𝑖,𝑥𝑖)) 11 end 12 cfs_min ← arg min (K𝑖,𝑥𝑖) ∈𝑐𝑎𝑛𝑑𝑖𝑑𝑎𝑡𝑒𝑠 𝛿(CBD(𝑥), CBD(𝑥𝑖)) 13 cfs_mean ← deepcopy(cfs_min) 14 Sort cfs_min by min-likeliness 15 Sort cfs_mean by mean-likeliness 16 return cfs_min, cfs_mean Algorithm 1: Generates two lists of counterfactuals of indi- vidual 𝑥 w.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='r.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='t.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' concept 𝐶—sorted by min-likeliness and mean- likeliness.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' with the least edit distance as defined in Equation 2 are selected as counterfactuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' All counterfactuals are then rated according to the two variants of likeliness (see Equations 3 and 4) and the algorithm returns two sorted lists of counterfactuals—one list for each measure.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 Analysis of the Generation of Counterfactuals In the following, we argue that every step carried out by the al- gorithms is necessary and that the steps carried out are sufficient.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' ELH allows concept intersection, existential restrictions and role hierarchy (see Table 1).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Because the concept 𝐶 is in ELH, we can assume that each 𝐶𝑖 (Algorithm 1, line 9) does not contain an in- tersection on the outer level anymore (such as 𝐶𝑖 = 𝐴 ⊓ 𝐵) and all remaining intersections must be within an existential restriction (such as 𝐶𝑖 = ∃𝑟.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' (𝐴 ⊓ 𝐵)).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' If the concept 𝐶 is an intersection, Al- gorithm 1 splits it into the subconcepts 𝐶𝑖 to apply Algorithm 2 to each subconcept 𝐶𝑖.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' If K ̸|= 𝐶𝑖 (𝑥) for any one 𝐶𝑖, it follows that K ̸|= 𝐶(𝑥) does not hold.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Creating additional candidates by making multiple 𝐶𝑖 not hold at once is not necessary, because these candidates could not exceed other candidates in minimizing the edit distance.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' If 𝐶𝑖 ∈ 𝑁𝐶, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' it is an atomic concept, all of the individual’s assertions to this concept or its subsumed subconcepts are removed (Algorithm 2, line 7).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Similarly, if 𝐶𝑖 is an existential restriction 𝐶𝑖 = ∃𝑟.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='𝐴, all role assertions 𝑟 ′(𝑥,𝑎) with K |= 𝐴(𝑎) and K |= 𝑟 ′ ⊑ 𝑟 are removed (Algorithm 2, line 9).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' This happens regardless of whether 𝐴 is a concept, another existential restriction, or an intersection.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' This is sufficient, because we aim only to remove axioms with the individual as subject to generate counterfactuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 1 Input: KB K, Concept 𝐶, Individual 𝑥 such that K |= 𝐶(𝑥) 2 Output: KB K, Individual 𝑥 such that K ̸|= 𝐶(𝑥) 3 Function gen_candidate(K, 𝐶, 𝑥): 4 if 𝐶 ≡ ⊤ then 5 K ← 𝑁𝑜𝑛𝑒 6 else if 𝐶 ∈ 𝑁𝐶 then 7 Remove assertions {𝐷(𝑥) | K |= 𝐷 ⊑ 𝐶 ∧ 𝐷 ∈ 𝑁𝐶} from K 8 else if 𝐶 = ∃𝑟.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='𝐴 then 9 Remove assertions {𝑟 ′(𝑥,𝑎) | K |= 𝐴(𝑎) ∧ K |= 𝑟 ′ ⊑ 𝑟} from K 10 return (K,𝑥) Algorithm 2: Generates a counterfactual candidate (K,𝑥) by changing the CBD of 𝑥 such that K ̸|= 𝐶(𝑥).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Table 2: Overview of the final, modified datasets in terms of number of instances (𝑁𝐼 ), axioms, atomic concepts and roles.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Instances Axioms Atomic Roles Dataset Concepts Family 202 2,033 18 5 Animals 28 170 19 4 5 EXPLANATIONS PREFERRED BY USERS We conducted a survey in which we let participants rate different potential counterfactual explanations against each other.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We then compared the participants’ preferences with the decisions made by our approach.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We used modified versions of the Family and Animals ontologies [37] to evaluate our approach.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' These ontologies were chosen because the concept, role and individual names therein are familiar and understandable to average lay users—in contrast to, for example, ontologies related to bio-medicine or chemistry.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We used the DL concept learner [8] with ELTL—the EL Tree Learner [7]—to learn the concepts to be used for counterfactual generation, since a future goal is to combine these programs to reach a fully automated explainable AI.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 Data Generation Table 2 gives an overview of the datasets used for our user survey.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 Family Ontology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We added a super role hasPartner of married to the original family ontology because the original on- tology had none and we wanted it to include a role hierarchy, so that we had an actual example of ELH and not just EL.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' To ob- tain (complex) concepts for the generation of counterfactuals, for each atomic concept present in the ontology, we did the following: First, we removed the atomic concept from the ontology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Then, we let the DL concept learner [8] with ELTL learn a concept using 10 randomly chosen individuals that formerly were instances of the atomic concept as positive examples, and 10 random others as negative examples.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Thereafter, we randomly selected an individual which is an instance of that concept and applied our counterfac- tual algorithm.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We manually inspected the learned concepts and used the correct concepts for the survey (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' concept Father leads Counterfactual Explanations for Concepts in ELH to Male ⊓ ∃hasChild.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='⊤).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For consistency, we also queried for the concepts for Brother and Grandmother as the corresponding con- cepts to Sister and Grandfather in the survey, even if ELTL did not correctly recognize the concept.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' This way, we ended up with Mother/Father, Sister/Brother and Grandmother/Grandfather for the survey.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2 Animals Ontology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We added two super roles to the animals ontology, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=', residence and home, for the same reasons as named above.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Furthermore, we restructured the ontology to fit ELH se- mantics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The atomic concepts of species were removed and their roles added to the individual animals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' This way we could train ELTL [8] to learn a concept for each species, using the instance be- longing to that species as positive example, and all other instances of animals as negative examples.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Our algorithm was applied after- wards.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We collected all concepts that could be used for explanations (see code at https://anonymous.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='4open.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='science/r/Counterfactual- Explanations-ELH-EBDA/ for details) and chose 6 of them randomly to use these and their counterfactual candidates for the survey.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2 Setup of User Survey The full survey material and data can be found at https://anonymous.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='4open.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='science/r/Counterfactual-Explanations- ELH-EBDA/.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Using the generated concepts from the family and animals ontologies (see above) and their respective counterfactual candidates, we generated short stories of artificial intelligences classifying people in a family tree or animals and created a counterfactual explanation from each counterfactual candidate.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We conducted an online survey via SoSciSurvey from May 1, 2022 to May 4, 2022 in German.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Participants were recruited through social networks and snowballing.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' On the first pages, participants were informed about the content and goal of the survey and what counterfactual explanations are, and we collected sociodemographic data about age, gender and occupation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' On the next pages, the counterfactual explanations were presented.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' First, a scenario was described in which an AI would classify instances of family members or animals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Then, on every page, a classification made by an AI was presented in one sentence, followed by one or multiple sentences giving counterfactual explanations for the classification, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' “I would not have classified this animal as a turtle, if it did not have scales”.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Within the two scenarios, classifications were presented in randomized order, one on each page.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For the family ontology, where each concept had led to two counterfactual candidates, the participants were randomly shown only one of the counterfactual explanations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' However, because many explanations were quite similar (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' all concepts included counterfactual candidates referring to gender) it was made sure that they were presented mixed explanation types.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Each explanation was accompanied by one item asking to rate on a scale from one to seven how helpful they perceived the explanation for understanding the decision of the program.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For the animals scenario, participants were shown all counterfactual candidates (between two and five) at the same time, in random order, and presented the same rating scale for each of the explanations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 Results of User Survey In the following, we present the results of our evaluation of our counterfactual rating algorithm through a user survey.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Sample.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 72 people took part in the survey.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Age ranged between 20 and 69 (average = 34.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='9, median = 32, standard deviation = 12.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1, missing age data for one participant).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 30 participants were female, 39 male and 3 diverse.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Participants had mixed professions including both academic and non-academic ones, technical and non-technical.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Family Ontology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We used Wilcoxon signed-rank tests to calcu- late significance of deviation from the central value (4 on a scale from 1-7) for each item and to compare both items of each concept with another.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Tests were chosen given the fact that we use ad-hoc generated items, so we assume the ratings to be ordinal.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Deviation results are presented in Table 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For all six concepts, users preferred the explanation that featured a role referring to relatives of the individual (hasChild, hasSibling) against the explanation that featured a 𝐶𝑖 referring to the individuals gender (all p<.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001, ex- cept Sister: p<.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='01).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' These explanations (and only these) differed significantly from the central value (all p<.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Comparison of user ratings with algorithmic decisions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For all the 4 individuals processed by our algorithms, the subconcepts 𝐶𝑖 ∈ 𝑁𝑅 were found to be counterfactuals by our algorithm, while the subconcepts 𝐶𝑖 ∈ 𝑁𝐶 were not.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Consequentially, both variants came to the same decision as the participants.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Animals ontology.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We used Wilcoxon signed-rank tests as above (but for matched samples).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For concepts with more than two coun- terfactuals, we used the Friedmann test.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 User evaluations of explanations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Table 3 presents results for Wilcoxon test of deviation from center and if and which algorithm variant chose this counterfactual.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' “Subconcept 𝐶𝑖” describes the subconcept that was used for explanation, e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' hasLegs means the counterfactual explanation contained “(.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='..' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=') if it did not have legs”.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2 Differences in user evaluations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For each animal, we calcu- lated which differences between the counterfactual explanations’ ratings were significant with a p value <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='05.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The difference be- tween the counterfactuals of Girl were not significant.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For Snake, habitat was rated significantly worse than the other counterfac- tuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For Penguin, only the difference between the best and worst counterfactual was significant.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For Eagle, the lowest rated two ex- planations (homeothermic and hasLegs) were rated significantly less helpful than the others.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Regarding Turtle, only the difference between the best (hasCovering(Scales)) and worst (hasLegs) ex- planation was significant.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Finally, for Crocodile, differences were not significant.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 Comparison of user ratings with algorithmic decisions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Al- though in the animals ontology, explanations similar to each other existed, too, participants’ decision about these varied.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The concept hasEggs was rather popular, while hasLegs was not.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Concepts habitat/residence were preferred 2 of 4 times.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For hasCovering, results were mixed;' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' concept homeothermic was never chosen.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We suspect that participants preferred explanations referring to fea- tures they find characteristic of the animal or groups of animals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For the animals ontology, all 𝐶𝑖 were counterfactuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Measuring Sieger et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Table 3: Ratings of counterfactual explanations (Wilcoxon signed-rank test).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The test statistics 𝜇, 𝑍, and 𝑝 refer to the user preferences of a subconcept 𝐶𝑖 of the named concept in the first column (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=', Mother ≡ Female ⊓ hasChild.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='⊤).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' ES stands for Effect Size.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Subconcepts in italic were chosen by algorithm with likeliness variant min.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Subconcepts in bold were chosen with variant mean.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The star(*) indicates p <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='01 and double-star(**) indicates p<.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Name Subconcept 𝐶𝑖 𝜇 Z p ES Mother Female 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='16 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='25 Mother hasChild(⊤) 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='83 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='84 Father Male 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='24 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='81 Father hasChild(⊤) 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='07 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='81 Sister Female 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='5 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='73 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='08 Sister hasSibling(⊤) 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='9 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='11 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='71 Brother Male 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='02 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='31 Brother hasSibling(⊤) 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='9 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='88 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='78 Grandmother Female 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='87 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='38 Grandmother hasChild(Parent) 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='88 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='78 Grandfather Male 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='79 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='43 Grandfather hasChild(Parent) 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='51 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='79 Girl HasLegs 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='67 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='09 Girl residence(⊤) 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='6 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='70 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='01* 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='32 Snake HasEggs 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='6 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='99 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='01* 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='35 Snake habitat(Land) 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='6 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='04 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='97 Snake hasCovering(Scales) 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='36 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='51 Penguin HasEggs 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='9 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='07 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='48 Penguin homeothermic 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='65 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='50 Penguin HasLegs 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='4 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='08 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='04 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='25 Penguin habitat(Water) 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='82 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='07 Penguin hasCovering(Feathers) 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='5 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='49 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='01 Eagle HasEggs 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='56 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='54 Eagle homeothermic 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='9 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='32 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='75 Eagle HasLegs 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='79 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='43 Eagle habitat(Air) 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='21 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='61 Eagle hasCovering(Feathers) 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='5 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='92 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='70 Turtle HasEggs 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='7 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='90 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='46 Turtle HasLegs 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='14 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='25 Turtle hasCovering(Scales) 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='6 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='04 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='96 Crocodile HasEggs 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='6 2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='87 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='01* 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='34 Crocodile HasLegs 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 1.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='50 .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='13 Crocodile hasCovering(Scales) 4.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='7 3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='42 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001** 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='40 the match of our algorithms with the participants’ ratings using F-score (2 · precision·recall precision+recall ), variantmin reached a score of 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3 while variantmean reached 0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='52.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Interpretation of results.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Overall, the survey results over both on- tologies can be interpreted as participants preferring explanations that contain features which are rather unlikely.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We assume that this is because such features are the most salient to distinguish between individuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In contrast, features which are very common (like be- ing of a certain gender or having legs) seem to be deemed less helpful.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' This fits our likeliness measurement idea, since removing a feature that does not appear often in the population for candidate generation should also result in a rather high likeliness score using our definitions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Our algorithm partly manages to cover that, but could be improved.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' variantmean seems to be more promising for that.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 6 STUDY ON USE CASES To assess possible use cases for counterfactual explanations in de- scription logics, we conducted an explorative study.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We wanted to know about users’ preferences for concept-based and/or coun- terfactual explanations, and how different use cases or con- cept length affect these.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The full study material and data can be found at https://anonymous.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='4open.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='science/r/Counterfactual- Explanations-ELH-EBDA/.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 Method We conducted an online survey via SoSciSurvey in German.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Par- ticipants were recruited from August 8, 2022 to August 15, 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Recruiting and setup of the study were similar to the survey.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' This time, participants were confronted with 4 different scenario stories.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The first two were similar to the ones in the survey (based on the family and animals ontologies), the others were fictional stories not based on existing ontologies.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' One was about an AI helping select the right medicine for a fictional person and in which case another medicament would be the better choice.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The other was about an AI deciding that the customer, called Clara, does not get a loan from the bank and what she could do to change this.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In this study, participants were both presented an explanation based on a concept (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' “I classify Petra as a mother, because she is female and has a child”) and a counterfactual explanation to rate.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For each scenario except the family scenario, participants were randomly assigned to one of two groups: For one, a long concept (size >3) was used for explanation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The second group was presented with shorter concepts.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Participants were asked to rate each information on a scale from 1-7 using three items: helpfulness for understanding the AI, usefulness, and if the information enhances controllability of the AI.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2 Results In this section, we present the results of our explorative study on explanation use cases.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 Sample.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 96 people took part in the survey.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Age ranged be- tween 20 and 70, mean = 35.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0, median = 32, standard deviation = 11.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 46 participants were female, 46 male, 3 diverse, one left the question blank.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 17 had professions related to IT, 33 worked in unrelated fields, 31 did not state it clearly enough to tell (e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='g.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=', student).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2 Factors related to explanation rating.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For each scenario and item, we used the Mann-Whitney U test to test for differences in concept-based explanation ratings from concept length.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' None of the results was significant (p>.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='01).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In one case, a difference between concept-based and counterfactual explanation was significant: In Counterfactual Explanations for Concepts in ELH the loan scenario, the counterfactual explanation was rated signifi- cantly higher on the item “This information is useful for Clara” than the long concept-based explanation (𝑍 = -3.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='29, 𝑝 = <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001, Effect size = .' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='34).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' To explore differences in ratings over scenarios, we used the Friedmann test.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We found that, again for the usefulness-item, the counterfactual explanation was significantly rated higher in the loan scenario than in the medicine or animal scenario (𝜒(3) = 25.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='10, 𝑝 <.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='0001) and also higher than in the family scenario.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' However, the latter difference was not significant by itself.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Concept ratings were not affected by scenario (Wilcoxon-signed rank test).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' All explana- tions had high ratings (average ratings ranging between 5.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='13 and 6.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='39).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 7 DISCUSSION We showed that our approach is able to generate counterfactuals with minimal edit distance measured by the distance of their concise bounded descriptions, i.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='e.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' few features changes to the individual.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Moreover, as there can be multiple counterfactuals per individual with minimal edit distance, we explored two likeliness measures to choose among them.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Regarding choosing the best counterfac- tual for an explanation, the results of our evaluation survey show room for improvement, but potential that an automated selection of explanations can be possible.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Concept-based vs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' counterfactual explanations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The explorative study gave us insight into people’s thoughts regarding concept- based versus counterfactual explanations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' While participants rated all explanations rather positively, we did not find much differences regarding the factors we took into account.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' However, regarding the loan scenario, we found that the counterfactual was rated more use- ful than the concept-based explanation.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The loan and the medicine scenario differed from the other two, that were already used in the first survey, in the fact that the AI’s classification directly affected a person.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' However, while the counterfactual was unclearly actionable (have a better cholesterol level) in the medicine scenario, the loan scenario’s counterfactual explanation was designed to provide a concrete possible action (pay off debts) to change the classifica- tion.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We suppose that this is the reason this explanation scored significantly higher on usefulness than the long concept and other counterfactuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Overall, our results suggest that counterfactuals perform at least as well as concept verbalization w.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='r.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='t.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' explaining algorithmic decisions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In addition, the study indicates that study- ing domains where counterfactual explanations lead to actionable decisions might be worthwhile.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1 Limitations and future work First of all, the target in our approach was to create a counterfac- tual that would not be classified as the chosen concept.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' However, in many applications, reaching a specific classification could be relevant.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Furthermore, we restricted the TBox to non-complex con- cept subsumptions, as such subsumptions rarely occur in practical ontologies.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Future work might relax this assumption.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The concepts we used were rather simple.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' A description logic that contains dis- junctions could be used for applications where explanations are more interesting to users, so we plan to expand our algorithm to the more complex DL ALC.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Structure of ontologies.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' A point open for discussion addresses different structures of ontologies.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' As it is the case in the family ontology, sometimes individuals might have or have not the same roles, while it is very unlikely for them to have the same objects for these roles, as argued before.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We therefore decided to just compare the sets of classes and roles of the individuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For other ontologies, this might not be a good decision.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' The same goes for the question if subfeatures of the changed feature should be counted into the edit distance, as we did here.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We will take a look at how various calcu- lation possibilities depend on use cases and ontology structures to be able to make specific recommendations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Actionability.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' While we tried to check for plausibility of coun- terfactual instances, our scoring mechanism cannot make sure yet that the axioms that were changed can actually be changed in the real world.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' One argument for counterfactual explanations is that in many applications it might be interesting for data subjects to get to know how they can change their classification [36].' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' However, the family ontology clearly shows an example of cases where this is not really possible, since people cannot usually change their gender or familiar relations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Poyiadzi et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [29] discuss the relevance of actionability of counterfactuals.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Since this is in line with our empir- ical findings, our future work will put more focus on applications where actionability can be reached and how to do this.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 7.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2 Applications The structure of the family ontology resembles data about individu- als in web sources such as DBpedia, YAGO, and Wikidata.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' To enrich such knowledge bases, additional information can be extracted from the web.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Concept learning allows checking the consistency of the extracted information and inferring new (implicit) axioms from the explicitly stated axioms.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Concept learning has been effectively applied to medical on- tologies [22], but the learned concepts can become very long [18], making them hard to grasp even for experts.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Counterfactuals, which might even be verbalized in natural language, help to steer focus to the most important parts of the concept.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Our study suggests that counterfactual explanations are most useful in actionable settings.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' For example, if a website is falsely classified as unsafe by a phishing site detector, the website owner might be interested in knowing which feature of his website caused the decision so that they can change it.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Ultimately, we want to develop a chatbot that, in the spirit of XAI, provides users with natural language explanations of automatically learned concepts and can be applied to various use cases in areas including web science, medicine and finance.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Meanwhile, we will look further into using counterfactual explanations for offering possible actions to alter classifications, to provide more agency to users.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 8 CONCLUSION We propose the first approach for generating counterfactual expla- nations in the description logic ELH.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Our approach performs well on the objective to generate counterfactual candidates for which a designated concept does not hold and which are similar to the individual.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' We discussed possibilities to improve the likeliness mea- surement of counterfactual candidates in accordance with findings Sieger et al.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' from a user study.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Our future work will move on to more complex DLs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' ACKNOWLEDGMENTS Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation): TRR 318/1 2021 – 438445824 REFERENCES [1] Amina Adadi and Mohammed Berrada.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Peeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' IEEE Access 6 (2018), 52138– 52160.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [2] Alejandro Barredo Arrieta, Natalia Díaz Rodríguez, Javier Del Ser, Adrien Ben- netot, Siham Tabik, Alberto Barbado, Salvador García, Sergio Gil-Lopez, Daniel Molina, Richard Benjamins, Raja Chatila, and Francisco Herrera.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Explain- able Artificial Intelligence (XAI): Concepts, taxonomies, opportunities and chal- lenges toward responsible AI.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Inf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Fusion 58 (2020), 82–115.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [3] Sören Auer, Christian Bizer, Georgi Kobilarov, Jens Lehmann, Richard Cyganiak, and Zachary Ives.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2007.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' DBpedia: A Nucleus for a Web of Open Data.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In The Semantic Web, Karl Aberer, Key-Sun Choi, Natasha Noy, Dean Allemang, Kyung- Il Lee, Lyndon Nixon, Jennifer Golbeck, Peter Mika, Diana Maynard, Riichiro Mizoguchi, Guus Schreiber, and Philippe Cudré-Mauroux (Eds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=').' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Springer Berlin Heidelberg, Berlin, Heidelberg, 722–735.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [4] Franz Baader, Ian Horrocks, Carsten Lutz, and Uli Sattler.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2017.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' An Introduction to Description Logic.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Cambridge University Press, Cambridge.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 1017/9781139025355 [5] Roi Blanco, Giuseppe Ottaviano, and Edgar Meij.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2015.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Fast and Space-Efficient Entity Linking for Queries.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Proceedings of the Eighth ACM International Con- ference on Web Search and Data Mining (Shanghai, China) (WSDM ’15).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' As- sociation for Computing Machinery, New York, NY, USA, 179–188.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https: //doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/2684822.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2685317 [6] S.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Brandt.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2004.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Reasoning in ELH w.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='r.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='t.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' General Concept Inclusion Axioms.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' LTCS- Report LTCS-04-03.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Chair for Automata Theory, Institute for Theoretical Com- puter Science, Dresden University of Technology, Germany.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' See http://lat.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='inf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='tu- dresden.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='de/research/reports.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='html.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='. [7] Lorenz Bühmann, Daniel Fleischhacker, Jens Lehmann, André Melo, and Johanna Völker.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2014.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Inductive Lexical Learning of Class Expressions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In EKAW (Lecture Notes in Computer Science, Vol.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 8876).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Springer, Cham, 42–53.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [8] Lorenz Bühmann, Jens Lehmann, and Patrick Westphal.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2016.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' DL-Learner—A framework for inductive learning on the Semantic Web.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Journal of Web Semantics 39 (2016), 15–24.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1016/j.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='websem.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2016.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='06.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='001 [9] Jiaoyan Chen, Xi Chen, Ian Horrocks, Erik B.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Myklebust, and Ernesto Jimenez- Ruiz.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Correcting Knowledge Base Assertions.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Proceedings of The Web Conference 2020 (Taipei, Taiwan) (WWW ’20).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Association for Computing Machin- ery, New York, NY, USA, 1537–1547.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/3366423.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3380226 [10] Susanne Dandl, Christoph Molnar, Martin Binder, and Bernd Bischl.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Multi- Objective Counterfactual Explanations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Parallel Problem Solving from Nature – PPSN XVI, Thomas Bäck, Mike Preuss, André Deutz, Hao Wang, Carola Do- err, Michael Emmerich, and Heike Trautmann (Eds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=').' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Springer International Publishing, Cham, 448–469.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [11] Caglar Demir and Axel-Cyrille Ngonga Ngomo.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' DRILL- Deep Reinforce- ment Learning for Refinement Operators in ALC.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' CoRR abs/2106.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='15373 (2021).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [12] Brigitte Grau and Anne-Laure Ligozat.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' A Corpus for Hybrid Question Answering Systems.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Companion Proceedings of the The Web Conference 2018 (Lyon, France) (WWW ’18).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' International World Wide Web Conferences Steering Committee, Republic and Canton of Geneva, CHE, 1081–1086.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/ 10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/3184558.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3191540 [13] Riccardo Guidotti, Anna Monreale, Salvatore Ruggieri, Franco Turini, Fosca Giannotti, and Dino Pedreschi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' A Survey of Methods for Explaining Black Box Models.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' ACM Comput.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Surv.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 51, 5 (2019), 93:1–93:42.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [14] Stefan Heindorf, Lukas Blübaum, Nick Düsterhus, Till Werner, Varun Nandku- mar Golani, Caglar Demir, and Axel-Cyrille Ngonga Ngomo.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' EvoLearner: Learning Description Logics with Evolutionary Algorithms.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Proceedings of the ACM Web Conference 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' ACM.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/3485447.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3511925 [15] Andre Hernich, Carsten Lutz, Ana Ozaki, and Frank Wolter.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2015.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Schema.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='Org as a Description Logic.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Proceedings of the 24th International Conference on Artificial Intelligence (IJCAI’15).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' AAAI Press, Buenos Aires, Argentina, 3048–3054.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [16] Aidan Hogan, Eva Blomqvist, Michael Cochez, Claudia d’Amato, Gerard de Melo, Claudio Gutiérrez, Sabrina Kirrane, José Emilio Labra Gayo, Roberto Navigli, Sebastian Neumaier, Axel-Cyrille Ngonga Ngomo, Axel Polleres, Sabbir M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Rashid, Anisa Rula, Lukas Schmelzeisen, Juan F.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Sequeda, Steffen Staab, and Antoine Zimmermann.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Knowledge Graphs.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Number 22 in Synthesis Lectures on Data, Semantics, and Knowledge.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Morgan & Claypool.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2200/ S01125ED1V01Y202109DSK022 [17] N’Dah Jean Kouagou, Stefan Heindorf, Caglar Demir, and Axel-Cyrille Ngonga Ngomo.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Neural Class Expression Synthesis.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='48550/ ARXIV.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2111.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='08486 [18] N’Dah Jean Kouagou, Stefan Heindorf, Caglar Demir, and Axel-Cyrille Ngonga Ngomo.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Learning Concept Lengths Accelerates Concept Learning in ALC.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In The Semantic Web, Paul Groth, Maria-Esther Vidal, Fabian Suchanek, Pedro Szekley, Pavan Kapanipathi, Catia Pesquita, Hala Skaf-Molli, and Minna Tamper (Eds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=').' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Springer International Publishing, Cham, 236–252.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [19] Maxat Kulmanov, Wang Liu-Wei, Yuan Yan, and Robert Hoehndorf.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' EL Embeddings: Geometric Construction of Models for the Description Logic EL++.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, IJCAI-19.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' International Joint Conferences on Artificial Intelligence Organization, 6103–6109.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='24963/ijcai.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2019/845 [20] Jonathan Lajus and Fabian M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Suchanek.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Are All People Married?' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Deter- mining Obligatory Attributes in Knowledge Bases.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Proceedings of the 2018 World Wide Web Conference (Lyon, France) (WWW ’18).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' International World Wide Web Conferences Steering Committee, Republic and Canton of Geneva, CHE, 1115–1124.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/3178876.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3186010 [21] Jens Lehmann.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2009.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' DL-Learner: Learning Concepts in Description Logics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Journal of Machine Learning Research 10 (2009), 2639–2642.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 1145/1577069.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1755874 [22] Jens Lehmann and Pascal Hitzler.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2010.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Concept learning in description logics using refinement operators.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Mach.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Learn.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 78, 1-2 (2010), 203–250.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [23] Karsten Lehmann and Anni-Yasmin Turhan.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2012.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' A Framework for Semantic- based Similarity Measures for ELH-Concepts.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Proceedings of the 13th Eu- ropean Conference on Logics in Artificial Intelligence (Lecture Notes in Artificial Intelligence), Luis Fariñas del Cerro, Andreas Herzig, and Jérôme Mengin (Eds.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' ).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Springer Verlag, 307–319.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [24] Jiaqi Li, Xuan Wu, Chang Lu, Wenxing Deng, and Yizheng Zhao.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Computing Views of OWL Ontologies for the Semantic Web.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Proceedings of the Web Conference 2021 (Ljubljana, Slovenia) (WWW ’21).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Association for Computing Machinery, New York, NY, USA, 2624–2635.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/3442381.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 3449881 [25] Christian Meske, Enrico Bunde, Johannes Schneider, and Martin Gersch.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Explainable artificial intelligence: objectives, stakeholders, and future research opportunities.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Information Systems Management 39, 1 (2022), 53–63.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [26] Christoph Molnar.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Interpretable Machine Learning: A Guide for Making Black Box Models Explainable.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' christophm.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='github.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='io/interpretable-ml-book/ [27] Carina Negreanu, Alperen Karaoglu, Jack Williams, Shuang Chen, Daniel Fabian, Andrew Gordon, and Chin-Yew Lin.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2022.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Rows from Many Sources: Enriching Row Completions from Wikidata with a Pre-Trained Language Model.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Compan- ion Proceedings of the Web Conference 2022 (Virtual Event, Lyon, France) (WWW ’22).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Association for Computing Machinery, New York, NY, USA, 1272–1280.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/3487553.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3524923 [28] Thomas Pellissier Tanon, Camille Bourgaux, and Fabian Suchanek.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Learning How to Correct a Knowledge Base from the Edit History.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In The World Wide Web Conference (San Francisco, CA, USA) (WWW ’19).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Association for Computing Machinery, New York, NY, USA, 1465–1475.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/3308558.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 3313584 [29] Rafael Poyiadzi, Kacper Sokol, Raul Santos-Rodriguez, Tijl De Bie, and Peter Flach.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' FACE: Feasible and Actionable Counterfactual Explanations.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society (New York, NY, USA) (AIES ’20).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Association for Computing Machinery, New York, NY, USA, 344–350.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/3375627.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3375850 [30] Charlotte Rudnik, Thibault Ehrhart, Olivier Ferret, Denis Teyssou, Raphael Troncy, and Xavier Tannier.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Searching News Articles Using an Event Knowledge Graph Leveraged by Wikidata.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Companion Proceedings of The 2019 World Wide Web Conference (San Francisco, USA) (WWW ’19).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Asso- ciation for Computing Machinery, New York, NY, USA, 1232–1239.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https: //doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/3308560.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3316761 [31] Md Kamruzzaman Sarker and Pascal Hitzler.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Efficient Concept Induction for Description Logics.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Proceedings of the AAAI Conference on Artificial Intelligence 33, 3036–3043.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1609/aaai.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='v33i01.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='33013036 [32] Ilia Stepin, Jose Alonso, Alejandro Catala, and Martin Pereira-Farina.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' A Survey of Contrastive and Counterfactual Explanation Generation Methods for Explainable Artificial Intelligence.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' IEEE Access 9 (01 2021), 11974–12001.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1109/ACCESS.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2021.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3051315 [33] Ilia Stepin, José Maria Alonso, Alejandro Catalá, and Martin Pereira-Fariña.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2020.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Generation and Evaluation of Factual and Gounterfaetual Explanations for Decision Trees and Fuzzy Rule-based Classifiers.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In FUZZ-IEEE.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' IEEE, 1–8.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' [34] Fabian M.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Suchanek, Gjergji Kasneci, and Gerhard Weikum.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2007.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Yago: A Core of Semantic Knowledge.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' In Proceedings of the 16th International Conference on World Wide Web (Banff, Alberta, Canada) (WWW ’07).' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Association for Computing Ma- chinery, New York, NY, USA, 697–706.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1145/1242572.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='1242667 [35] Denny Vrandečić and Markus Krötzsch.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2014.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Wikidata: A Free Collaborative Knowledgebase.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Commun.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' ACM 57, 10 (sep 2014), 78–85.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https://doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 1145/2629489 Counterfactual Explanations for Concepts in ELH [36] Sandra Wachter, Brent Mittelstadt, and Chris Russell.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2018.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Counterfactual Explanations Without Opening the Black Box: Automated Decisions and the GDPR.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Harvard journal of law & technology 31 (04 2018), 841–887.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' https: //doi.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='org/10.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='2139/ssrn.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content='3063289 [37] Patrick Westphal, Lorenz Bühmann, Simon Bin, Hajira Jabeen, and Jens Lehmann.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' 2019.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' SML-Bench - A benchmarking framework for structured machine learning.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'} +page_content=' Semantic Web 10, 2 (2019), 231–245.' metadata={'source': '/home/zjlab/wf/langchain-ChatGLM/knowledge_base/FNE4T4oBgHgl3EQffg3m/content/2301.05109v1.pdf'}