xml
stringlengths 7
23
| proceedings
stringlengths 58
222
| year
stringdate 2005-01-01 00:00:00
2024-01-01 00:00:00
| url
stringlengths 1
64
| language documentation
stringclasses 1
value | has non-English?
stringclasses 4
values | topics
stringclasses 7
values | language coverage
stringclasses 32
values | title
stringlengths 32
161
| abstract
stringlengths 176
2.45k
|
|---|---|---|---|---|---|---|---|---|---|
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.115 | x | 0 | toxicity, bias | null | CMD: a framework for Context-aware Model self-Detoxification | Text detoxification aims to minimize the risk of language models producing toxic content. Existing detoxification methods of directly constraining the model output or further training the model on the non-toxic corpus fail to achieve a decent balance between detoxification effectiveness and generation quality. This issue stems from the neglect of constrain imposed by the context since language models are designed to generate output that closely matches the context while detoxification methods endeavor to ensure the safety of the output even if it semantically deviates from the context. In view of this, we introduce a Context-aware Model self-Detoxification (CMD) framework that pays attention to both the context and the detoxification process, i.e., first detoxifying the context and then making the language model generate along the safe context. Specifically, CMD framework involves two phases: utilizing language models to synthesize data and applying these data for training. We also introduce a toxic contrastive loss that encourages the model generation away from the negative toxic samples. Experiments on various LLMs have verified the effectiveness of our MSD framework, which can yield the best performance compared to baselines. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.157 | null | 0 | jailbreaking attacks | null | ASETF: A Novel Method for Jailbreak Attack on LLMs through Translate Suffix Embeddings | The safety defense methods of Large language models (LLMs) stays limited because the dangerous prompts are manually curated to just few known attack types, which fails to keep pace with emerging varieties. Recent studies found that attaching suffixes to harmful instructions can hack the defense of LLMs and lead to dangerous outputs. However, similar to traditional text adversarial attacks, this approach, while effective, is limited by the challenge of the discrete tokens. This gradient based discrete optimization attack requires over 100,000 LLM calls, and due to the unreadable of adversarial suffixes, it can be relatively easily penetrated by common defense methods such as perplexity filters.To cope with this challenge, in this paper, we propose an Adversarial Suffix Embedding Translation Framework (ASETF), aimed at transforming continuous adversarial suffix embeddings into coherent and understandable text. This method greatly reduces the computational overhead during the attack process and helps to automatically generate multiple adversarial samples, which can be used as data to strengthen LLM’s security defense. Experimental evaluations were conducted on Llama2, Vicuna, and other prominent LLMs, employing harmful directives sourced from the Advbench dataset.The results indicate that our method significantly reduces the computation time of adversarial suffixes and achieves a much better attack success rate than existing techniques, while significantly enhancing the textual fluency of the prompts. In addition, our approach can be generalized into a broader method for generating transferable adversarial suffixes that can successfully attack multiple LLMs, even black-box LLMs, such as ChatGPT and Gemini. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.164 | x | 0 | jailbreaking attacks | null | Alignment-Enhanced Decoding: Defending Jailbreaks via Token-Level Adaptive Refining of Probability Distributions | Large language models are susceptible to jailbreak attacks, which can result in the generation of harmful content. While prior defenses mitigate these risks by perturbing or inspecting inputs, they ignore competing objectives, the underlying cause of alignment failures. In this paper, we propose Alignment-Enhanced Decoding (AED), a novel defense that employs adaptive decoding to address the root causes of jailbreak issues. We first define the Competitive Index to quantify alignment failures and utilize feedback from self-evaluation to compute post-alignment logits. Then, AED adaptively combines Competitive Index and post-alignment logits with the original logits to obtain harmless and helpful distributions. Consequently, our method enhances safety alignment while maintaining helpfulness. We conduct experiments across five models and four common jailbreaks, with the results validating the effectiveness of our approach. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.201 | null | 0 | general safety, LLM alignment | null | Is Safer Better? The Impact of Guardrails on the Argumentative Strength of LLMs in Hate Speech Countering | The potential effectiveness of counterspeech as a hate speech mitigation strategy is attracting increasing interest in the NLG research community, particularly towards the task of automatically producing it. However, automatically generated responses often lack the argumentative richness which characterises expert-produced counterspeech. In this work, we focus on two aspects of counterspeech generation to produce more cogent responses. First, by investigating the tension between helpfulness and harmlessness of LLMs, we test whether the presence of safety guardrails hinders the quality of the generations. Secondly, we assess whether attacking a specific component of the hate speech results in a more effective argumentative strategy to fight online hate. By conducting an extensive human and automatic evaluation, we show how the presence of safety guardrails can be detrimental also to a task that inherently aims at fostering positive social interactions. Moreover, our results show that attacking a specific component of the hate speech, and in particular its implicit negative stereotype and its hateful parts, leads to higher-quality generations. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.277 | null | 0 | others | null | KidLM: Advancing Language Models for Children – Early Insights and Future Directions | Recent studies highlight the potential of large language models in creating educational tools for children, yet significant challenges remain in maintaining key child-specific properties such as linguistic nuances, cognitive needs, and safety standards. In this paper, we explore foundational steps toward the development of child-specific language models, emphasizing the necessity of high-quality pre-training data. We introduce a novel user-centric data collection pipeline that involves gathering and validating a corpus specifically written for and sometimes by children. Additionally, we propose a new training objective, Stratified Masking, which dynamically adjusts masking probabilities based on our domain-specific child language data, enabling models to prioritize vocabulary and concepts more suitable for children. Experimental evaluations demonstrate that our model excels in understanding lower grade-level text, maintains safety by avoiding stereotypes, and captures children’s unique preferences. Furthermore, we provide actionable insights for future research and development in child-specific language modeling. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.280 | x | 0 | others | null | An Unsupervised Approach to Achieve Supervised-Level Explainability in Healthcare Records | Electronic healthcare records are vital for patient safety as they document conditions, plans, and procedures in both free text and medical codes. Language models have significantly enhanced the processing of such records, streamlining workflows and reducing manual data entry, thereby saving healthcare providers significant resources. However, the black-box nature of these models often leaves healthcare professionals hesitant to trust them. State-of-the-art explainability methods increase model transparency but rely on human-annotated evidence spans, which are costly. In this study, we propose an approach to produce plausible and faithful explanations without needing such annotations. We demonstrate on the automated medical coding task that adversarial robustness training improves explanation plausibility and introduce AttInGrad, a new explanation method superior to previous ones. By combining both contributions in a fully unsupervised setup, we produce explanations of comparable quality, or better, to that of a supervised approach. We release our code and model weights. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.358 | null | 0 | hallucination, factuality | null | VLFeedback: A Large-Scale AI Feedback Dataset for Large Vision-Language Models Alignment | As large vision-language models (LVLMs) evolve rapidly, the demand for high-quality and diverse data to align these models becomes increasingly crucial. However, the creation of such data with human supervision proves costly and time-intensive. In this paper, we investigate the efficacy of AI feedback to scale supervision for aligning LVLMs. We introduce VLFeedback, the first large-scale vision-language feedback dataset, comprising over 82K multi-modal instructions and comprehensive rationales generated by off-the-shelf models without human annotations. To evaluate the effectiveness of AI feedback for vision-language alignment, we train Silkie, an LVLM fine-tuned via direct preference optimization on VLFeedback. Silkie showcases exceptional performance regarding helpfulness, visual faithfulness, and safety metrics. It outperforms its base model by 6.9% and 9.5% in perception and cognition tasks, reduces hallucination issues on MMHal-Bench, and exhibits enhanced resilience against red-teaming attacks. Furthermore, our analysis underscores the advantage of AI feedback, particularly in fostering preference diversity to deliver more comprehensive improvements. Our dataset, training code and models are available at https://vlf-silkie.github.io. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.409 | x | 0 | jailbreaking attacks | null | Advancing Adversarial Suffix Transfer Learning on Aligned Large Language Models | Language Language Models (LLMs) face safety concerns due to potential misuse by malicious users. Recent red-teaming efforts have identified adversarial suffixes capable of jailbreaking LLMs using the gradient-based search algorithm Greedy Coordinate Gradient (GCG). However, GCG struggles with computational inefficiency, limiting further investigations regarding suffix transferability and scalability across models and data. In this work, we bridge the connection between search efficiency and suffix transferability. We propose a two-stage transfer learning framework, DeGCG, which decouples the search process into behavior-agnostic pre-searching and behavior-relevant post-searching. Specifically, we employ direct first target token optimization in pre-searching to facilitate the search process. We apply our approach to cross-model, cross-data, and self-transfer scenarios. Furthermore, we introduce an interleaved variant of our approach, i-DeGCG, which iteratively leverages self-transferability to accelerate the search process. Experiments on HarmBench demonstrate the efficiency of our approach across various models and domains. Notably, our i-DeGCG outperforms the baseline on Llama2-chat-7b with ASRs of 43.9 (+ 22.2) and 39.0 (+19.5) on valid and test sets, respectively. Further analysis on cross-model transfer indicates the pivotal role of first target token optimization in leveraging suffix transferability for efficient searching. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.445 | x | 0 | toxicity, bias | null | Bridging Modalities: Enhancing Cross-Modality Hate Speech Detection with Few-Shot In-Context Learning | The widespread presence of hate speech on the internet, including formats such as text-based tweets and multimodal memes, poses a significant challenge to digital platform safety. Recent research has developed detection models tailored to specific modalities; however, there is a notable gap in transferring detection capabilities across different formats. This study conducts extensive experiments using few-shot in-context learning with large language models to explore the transferability of hate speech detection between modalities. Our findings demonstrate that text-based hate speech examples can significantly enhance the classification accuracy of vision-language hate speech. Moreover, text-based demonstrations outperform vision-language demonstrations in few-shot learning settings. These results highlight the effectiveness of cross-modality knowledge transfer and offer valuable insights for improving hate speech detection systems. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.461 | x | 0 | general safety, LLM alignment | null | Data Advisor: Dynamic Data Curation for Safety Alignment of Large Language Models | Data are crucial element in large language model (LLM) alignment. Recent studies have explored using LLMs for efficient data collection. However, LLM-generated data often suffers from quality issues, with underrepresented or absent aspects and low-quality datapoints. To address these problems, we propose Data Advisor, an enhanced LLM-based method for generating data that takes into account the characteristics of the desired dataset. Starting from a set of pre-defined principles in hand, Data Advisor monitors the status of the generated data, identifies weaknesses in the current dataset, and advises the next iteration of data generation accordingly. Data Advisor can be easily integrated into existing data generation methods to enhance data quality and coverage. Experiments on safety alignment of three representative LLMs (i.e., Mistral, Llama2, and Falcon) demonstrate the effectiveness of Data Advisor in enhancing model safety against various fine-grained safety issues without sacrificing model utility. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.511 | null | 0 | general safety, LLM alignment | null | Annotation alignment: Comparing LLM and human annotations of conversational safety | Do LLMs align with human perceptions of safety? We study this question via *annotation alignment*, the extent to which LLMs and humans agree when annotating the safety of user-chatbot conversations. We leverage the recent DICES dataset (Aroyo et al. 2023), in which 350 conversations are each rated for safety by 112 annotators spanning 10 race-gender groups. GPT-4 achieves a Pearson correlation of r=0.59 with the average annotator rating, higher than the median annotator’s correlation with the average (r=0.51). We show that larger datasets are needed to resolve whether GPT-4 exhibits disparities in how well it correlates with different demographic groups. Also, there is substantial idiosyncratic variation in correlation within groups, suggesting that race & gender do not fully capture differences in alignment. Finally, we find that GPT-4 cannot predict when one demographic group finds a conversation more unsafe than another. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.534 | x | 0 | jailbreaking attacks | null | Ranking Manipulation for Conversational Search Engines | Major search engine providers are rapidly incorporating Large Language Model (LLM)-generated content in response to user queries. These *conversational search engines* operate by loading retrieved website text into the LLM context for summarization and interpretation. Recent research demonstrates that LLMs are highly vulnerable to jailbreaking and prompt injection attacks, which disrupt the safety and quality goals of LLMs using adversarial strings. This work investigates the impact of prompt injections on the ranking order of sources referenced by conversational search engines. To this end, we introduce a focused dataset of real-world consumer product websites and formalize conversational search ranking as an adversarial problem. Experimentally, we analyze conversational search rankings in the absence of adversarial injections and show that different LLMs vary significantly in prioritizing product name, document content, and context position. We then present a tree-of-attacks-based jailbreaking technique which reliably promotes low-ranked products. Importantly, these attacks transfer effectively to state-of-the-art conversational search engines such as *perplexity.ai*. Given the strong financial incentive for website owners to boost their search ranking, we argue that our problem formulation is of critical importance for future robustness work. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.585 | x | 0 | general safety, LLM alignment | null | InferAligner: Inference-Time Alignment for Harmlessness through Cross-Model Guidance | As large language models (LLMs) rapidly evolve, they are increasingly being customized through fine-tuning to suit the specific needs of various applications. A critical aspect of this advancement is the alignment process, which ensures that these models perform tasks in ways that align with human values and expectations. Current alignment methods, such as direct preference optimization (DPO) and reinforcement learning from human feedback (RLHF), focus primarily on alignment during training phase. However, these methods often involve complex and resource-intensive training processes, posing significant challenge for their implementation. Therefore, we propose InferAligner, a simple yet effective method for harmlessness alignment during inference phase. InferAligner decouples harmlessness from helpfulness. During the training phase, it focuses solely on enhancing the target model’s capabilities on downstream tasks. In the inference phase, it utilizes safety steering vectors extracted from the aligned model to guide the target model towards harmlessness alignment. Experimental results show that our method can be very effectively applied to domain-specific models in finance, medicine, and mathematics, as well as to multimodal large language models (MLLMs) such as LLaVA. It significantly diminishes the attack success rate (ASR) of both harmful instructions and jailbreak instructions, while maintaining almost unchanged performance in downstream tasks. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.649 | null | 0 | others | null | Fishing for Magikarp: Automatically Detecting Under-trained Tokens in Large Language Models | The disconnect between tokenizer creation and model training in language models allows for specific inputs, such as the infamous SolidGoldMagikarp token, to induce unwanted model behaviour. Although such ‘glitch tokens’, tokens present in the tokenizer vocabulary but that are nearly or entirely absent during model training, have been observed across various models, a reliable method to identify and address them has been missing. We present a comprehensive analysis of Large Language Model tokenizers, specifically targeting this issue of detecting under-trained tokens. Through a combination of tokenizer analysis, model weight-based indicators, and prompting techniques, we develop novel and effective methods for automatically detecting these problematic tokens. Our findings demonstrate the prevalence of such tokens across a diverse set of models and provide insights into improving the efficiency and safety of language models. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.656 | x | 0 | general safety, LLM alignment | null | LoRA-Guard: Parameter-Efficient Guardrail Adaptation for Content Moderation of Large Language Models | Guardrails have emerged as an alternative to safety alignment for content moderation of large language models (LLMs). Existing model-based guardrails have not been designed for resource-constrained computational portable devices, such as mobile phones, more and more of which are running LLM-based applications locally. We introduce LoRA-Guard, a parameter-efficient guardrail adaptation method that relies on knowledge sharing between LLMs and guardrail models. LoRA-Guard extracts language features from the LLMs and adapts them for the content moderation task using low-rank adapters, while a dual-path design prevents any performance degradation on the generative task. We show that LoRA-Guard outperforms existing approaches with 100-1000x lower parameter overhead while maintaining accuracy, enabling on-device content moderation. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.671 | null | 1 | general safety, LLM alignment | English, French, Spanish, Hindi, Arabic, Russian | The Multilingual Alignment Prism: Aligning Global and Local Preferences to Reduce Harm | A key concern with the concept of *“alignment”* is the implicit question of *“alignment to what?”*. AI systems are increasingly used across the world, yet safety alignment is often focused on homogeneous monolingual settings. Additionally, preference training and safety measures often overfit to harms common in Western-centric datasets. Here, we explore the viability of different alignment approaches when balancing dual objectives: addressing and optimizing for a non-homogeneous set of languages and cultural preferences while minimizing both global and local harms. We collect the first human annotated red teaming prompts in different languages, distinguishing between global and local harm, which serve as a laboratory to understand the reliability of alignment techniques when faced with preference distributions that are non-stationary across geographies and languages. While this setting is seldom covered by the literature to date, which primarily centers on English harm mitigation, it captures real-world interactions with AI systems around the world. We establish a new precedent for state-of-the-art alignment techniques across 6 languages with minimal degradation in general performance. Our work provides important insights into cross-lingual transfer and novel optimization approaches to safeguard AI systems designed to serve global populations. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.732 | x | 0 | jailbreaking attacks | null | BEEAR: Embedding-based Adversarial Removal of Safety Backdoors in Instruction-tuned Language Models | Safety backdoor attacks in large language models (LLMs) enable harmful behaviors to be stealthily triggered while evading detection during normal interactions. The high dimensionality of the trigger search space and the diverse range of potential malicious behaviors in LLMs make this a critical open problem. This paper presents BEEAR, a novel mitigation method based on a key insight: backdoor triggers induce a uniform drift in the model’s embedding space, irrespective of the trigger’s form or targeted behavior. Leveraging this observation, we introduce a bi-level optimization approach. The inner level identifies universal perturbations to the decoder’s embeddings that steer the model towards defender-defined unwanted behaviors; the outer level fine-tunes the model to reinforce safe behaviors against these perturbations. Our experiments demonstrate the effectiveness of this approach, reducing the success rate of safety backdoor attacks from over 95% to <1% for general harmful behaviors and from 47% to 0% for Sleeper Agents, without compromising the model’s helpfulness. Notably, our method relies only on defender-defined sets of safe and unwanted behaviors without any assumptions about the trigger location or attack mechanism. This work represents the first practical framework to counter safety backdoors in LLMs and provides a foundation for future advancements in AI safety and security. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.738 | x | 0 | jailbreaking attacks | null | Large Language Models Are Involuntary Truth-Tellers: Exploiting Fallacy Failure for Jailbreak Attacks | We find that language models have difficulties generating fallacious and deceptive reasoning. When asked to generate deceptive outputs, language models tend to leak honest counterparts but believe them to be false. Exploiting this deficiency, we propose a jailbreak attack method that elicits an aligned language model for malicious output. Specifically, we query the model to generate a fallacious yet deceptively real procedure for the harmful behavior. Since a fallacious procedure is generally considered fake and thus harmless by LLMs, it helps bypass the safeguard mechanism. Yet the output is factually harmful since the LLM cannot fabricate fallacious solutions but proposes truthful ones. We evaluate our approach over five safety-aligned large language models, comparing four previous jailbreak methods, and show that our approach achieves competitive performance with more harmful outputs. We believe the findings could be extended beyond model safety, such as self-verification and hallucination. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.761 | null | 0 | others | null | Householder Pseudo-Rotation: A Novel Approach to Activation Editing in LLMs with Direction-Magnitude Perspective | Activation Editing, which involves directly editting the internal representations of large language models (LLMs) to alter their behavior and achieve desired properties, has emerged as a promising area of research. Existing works primarily treat LLMs’ activations as points in space and modify them by adding steering vectors. We show that doing so would break the magnitude consistency of the activation vectors in LLMs. To overcome this shortcoming, we propose a novel editing method that views activations in terms of their directions and magnitudes. Our method, which we name Householder Pseudo-Rotation (HPR), mimics the rotation transformation, thus preserving activation norm and resulting in an improved performance on various safety benchmarks. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.806 | x | 0 | others | null | SecCoder: Towards Generalizable and Robust Secure Code Generation | After large models (LMs) have gained widespread acceptance in code-related tasks, their superior generative capacity has greatly promoted the application of the code LM. Nevertheless, the security of the generated code has raised attention to its potential damage. Existing secure code generation methods have limited generalizability to unseen test cases and poor robustness against the attacked model, leading to safety failures in code generation. In this paper, we propose a generalizable and robust secure code generation method SecCoder by using in-context learning (ICL) and the safe demonstration. The dense retriever is also used to select the most helpful demonstration to maximize the improvement of the generated code’s security. Experimental results show the superior generalizability of the proposed model SecCoder compared to the current secure code generation method, achieving a significant security improvement of an average of 7.20% on unseen test cases. The results also show the better robustness of SecCoder compared to the current attacked code LM, achieving a significant security improvement of an average of 7.74%. Our analysis indicates that SecCoder enhances the security of LMs in generating code, and it is more generalizable and robust. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.895 | null | 0 | others | null | MLLM-Protector: Ensuring MLLM’s Safety without Hurting Performance | The deployment of multimodal large language models (MLLMs) has brought forth a unique vulnerability: susceptibility to malicious attacks through visual inputs. This paper investigates the novel challenge of defending MLLMs against such attacks. Compared to large language models (LLMs), MLLMs include an additional image modality. We discover that images act as a “foreign language” that is not considered during safety alignment, making MLLMs more prone to producing harmful responses. Unfortunately, unlike the discrete tokens considered in text-based LLMs, the continuous nature of image signals presents significant alignment challenges, which poses difficulty to thoroughly cover all possible scenarios. This vulnerability is exacerbated by the fact that most state-of-the-art MLLMs are fine-tuned on limited image-text pairs that are much fewer than the extensive text-based pretraining corpus, which makes the MLLMs more prone to catastrophic forgetting of their original abilities during safety fine-tuning. To tackle these challenges, we introduce MLLM-Protector, a plug-and-play strategy that solves two subtasks: 1) identifying harmful responses via a lightweight harm detector, and 2) transforming harmful responses into harmless ones via a detoxifier. This approach effectively mitigates the risks posed by malicious visual inputs without compromising the original performance of MLLMs. Our results demonstrate that MLLM-Protector offers a robust solution to a previously unaddressed aspect of MLLM security. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.968 | null | 0 | jailbreaking attacks | null | CoSafe: Evaluating Large Language Model Safety in Multi-Turn Dialogue Coreference | As large language models (LLMs) constantly evolve, ensuring their safety remains a critical research issue. Previous red teaming approaches for LLM safety have primarily focused on single prompt attacks or goal hijacking. To the best of our knowledge, we are the first to study LLM safety in multi-turn dialogue coreference. We created a dataset of 1,400 questions across 14 categories, each featuring multi-turn coreference safety attacks. We then conducted detailed evaluations on five widely used open-source LLMs. The results indicated that under multi-turn coreference safety attacks, the highest attack success rate was 56% with the LLaMA2-Chat-7b model, while the lowest was 13.9% with the Mistral-7B-Instruct model. These findings highlight the safety vulnerabilities in LLMs during dialogue coreference interactions. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.1022 | null | 1 | general safety, LLM alignment | German, French, Italian, Spanish | GuardBench: A Large-Scale Benchmark for Guardrail Models | Generative AI systems powered by Large Language Models have become increasingly popular in recent years. Lately, due to the risk of providing users with unsafe information, the adoption of those systems in safety-critical domains has raised significant concerns. To respond to this situation, input-output filters, commonly called guardrail models, have been proposed to complement other measures, such as model alignment. Unfortunately, the lack of a standard benchmark for guardrail models poses significant evaluation issues and makes it hard to compare results across scientific publications. To fill this gap, we introduce GuardBench, a large-scale benchmark for guardrail models comprising 40 safety evaluation datasets. To facilitate the adoption of GuardBench, we release a Python library providing an automated evaluation pipeline built on top of it. With our benchmark, we also share the first large-scale prompt moderation datasets in German, French, Italian, and Spanish. To assess the current state-of-the-art, we conduct an extensive comparison of recent guardrail models and show that a general-purpose instruction-following model of comparable size achieves competitive results without the need for specific fine-tuning. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.1033 | null | 0 | hallucination, factuality | null | Enhanced Hallucination Detection in Neural Machine Translation through Simple Detector Aggregation | Hallucinated translations pose significant threats and safety concerns when it comes to practical deployment of machine translation systems. Previous research works have identified that detectors exhibit complementary performance — different detectors excel at detecting different types of hallucinations. In this paper, we propose to address the limitations of individual detectors by combining them and introducing a straightforward method for aggregating multiple detectors. Our results demonstrate the efficacy of our aggregated detector, providing a promising step towards evermore reliable machine translation systems. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.1034 | null | 1 | jailbreaking attacks | Arabic | Jailbreaking LLMs with Arabic Transliteration and Arabizi | This study identifies the potential vulnerabilities of Large Language Models (LLMs) to ‘jailbreak’ attacks, specifically focusing on the Arabic language and its various forms. While most research has concentrated on English-based prompt manipulation, our investigation broadens the scope to investigate the Arabic language. We initially tested the AdvBench benchmark in Standardized Arabic, finding that even with prompt manipulation techniques like prefix injection, it was insufficient to provoke LLMs into generating unsafe content. However, when using Arabic transliteration and chatspeak (or arabizi), we found that unsafe content could be produced on platforms like OpenAI GPT-4 and Anthropic Claude 3 Sonnet. Our findings suggest that using Arabic and its various forms could expose information that might remain hidden, potentially increasing the risk of jailbreak attacks. We hypothesize that this exposure could be due to the model’s learned connection to specific words, highlighting the need for more comprehensive safety training across all language forms. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.1049 | null | 1 | hallucination, factuality | English, German, Spanish, Hindi | Are Data Augmentation Methods in Named Entity Recognition Applicable for Uncertainty Estimation? | This work investigates the impact of data augmentation on confidence calibration and uncertainty estimation in Named Entity Recognition (NER) tasks. For the future advance of NER in safety-critical fields like healthcare and finance, it is essential to achieve accurate predictions with calibrated confidence when applying Deep Neural Networks (DNNs), including Pre-trained Language Models (PLMs), as a real-world application. However, DNNs are prone to miscalibration, which limits their applicability. Moreover, existing methods for calibration and uncertainty estimation are computational expensive. Our investigation in NER found that data augmentation improves calibration and uncertainty in cross-genre and cross-lingual setting, especially in-domain setting. Furthermore, we showed that the calibration for NER tends to be more effective when the perplexity of the sentences generated by data augmentation is lower, and that increasing the size of the augmentation further improves calibration and uncertainty. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.1051 | x | 0 | others | null | Fool Me Once? Contrasting Textual and Visual Explanations in a Clinical Decision-Support Setting | The growing capabilities of AI models are leading to their wider use, including in safety-critical domains. Explainable AI (XAI) aims to make these models safer to use by making their inference process more transparent. However, current explainability methods are seldom evaluated in the way they are intended to be used: by real-world end users. To address this, we conducted a large-scale user study with 85 healthcare practitioners in the context of human-AI collaborative chest X-ray analysis. We evaluated three types of explanations: visual explanations (saliency maps), natural language explanations, and a combination of both modalities. We specifically examined how different explanation types influence users depending on whether the AI advice and explanations are factually correct. We find that text-based explanations lead to significant over-reliance, which is alleviated by combining them with saliency maps. We also observe that the quality of explanations, that is, how much factually correct information they entail, and how much this aligns with AI correctness, significantly impacts the usefulness of the different explanation types. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.1090 | null | 0 | toxicity, bias | null | Adaptable Moral Stances of Large Language Models on Sexist Content: Implications for Society and Gender Discourse | This work provides an explanatory view of how LLMs can apply moral reasoning to both criticize and defend sexist language. We assessed eight large language models, all of which demonstrated the capability to provide explanations grounded in varying moral perspectives for both critiquing and endorsing views that reflect sexist assumptions. With both human and automatic evaluation, we show that all eight models produce comprehensible and contextually relevant text, which is helpful in understanding diverse views on how sexism is perceived. Also, through analysis of moral foundations cited by LLMs in their arguments, we uncover the diverse ideological perspectives in models’ outputs, with some models aligning more with progressive or conservative views on gender roles and sexism.Based on our observations, we caution against the potential misuse of LLMs to justify sexist language. We also highlight that LLMs can serve as tools for understanding the roots of sexist beliefs and designing well-informed interventions. Given this dual capacity, it is crucial to monitor LLMs and design safety mechanisms for their use in applications that involve sensitive societal topics, such as sexism. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.1200 | null | 0 | general safety, LLM alignment | null | STAR: SocioTechnical Approach to Red Teaming Language Models | This research introduces STAR, a sociotechnical framework that improves on current best practices for red teaming safety of large language models. STAR makes two key contributions: it enhances steerability by generating parameterised instructions for human red teamers, leading to improved coverage of the risk surface. Parameterised instructions also provide more detailed insights into model failures at no increased cost. Second, STAR improves signal quality by matching demographics to assess harms for specific groups, resulting in more sensitive annotations. STAR further employs a novel step of arbitration to leverage diverse viewpoints and improve label reliability, treating disagreement not as noise but as a valuable contribution to signal quality. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.1212 | x | 0 | general safety, LLM alignment | null | Safety Arithmetic: A Framework for Test-time Safety Alignment of Language Models by Steering Parameters and Activations | Ensuring the safe alignment of large language models (LLMs) with human values is critical as they become integral to applications like translation and question answering. Current alignment methods struggle with dynamic user intentions and complex objectives, making models vulnerable to generating harmful content. We propose Safety Arithmetic, a training-free framework enhancing LLM safety across different scenarios: Base models, Supervised fine-tuned models (SFT), and Edited models. Safety Arithmetic involves Harm Direction Removal to avoid harmful content and Safety Alignment to promote safe responses. Additionally, we present NoIntentEdit, a dataset highlighting edit instances that could compromise model safety if used unintentionally. Our experiments show that Safety Arithmetic significantly improves safety measures, reduces over-safety, and maintains model utility, outperforming existing methods in ensuring safe content generation. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-main.1235 | x | 0 | jailbreaking attacks | null | GPT-4 Jailbreaks Itself with Near-Perfect Success Using Self-Explanation | Research on jailbreaking has been valuable for testing and understanding the safety and security issues of large language models (LLMs). In this paper, we introduce Iterative Refinement Induced Self-Jailbreak (IRIS), a novel approach that leverages the reflective capabilities of LLMs for jailbreaking with only black-box access. Unlike previous methods, IRIS simplifies the jailbreaking process by using a single model as both the attacker and target. This method first iteratively refines adversarial prompts through self-explanation, which is crucial for ensuring that even well-aligned LLMs obey adversarial instructions. IRIS then rates and enhances the output given the refined prompt to increase its harmfulness. We find that IRIS achieves jailbreak success rates of 98% on GPT-4, 92% on GPT-4 Turbo, and 94% on Llama-3.1-70B in under 7 queries. It significantly outperforms prior approaches in automatic, black-box, and interpretable jailbreaking, while requiring substantially fewer queries, thereby establishing a new standard for interpretable jailbreaking methods. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-demo.20 | x | 0 | others | null | WebOlympus: An Open Platform for Web Agents on Live Websites | Web agents are emerging as powerful tools capable of performing complex tasks across diverse web environments. The rapid development of large multimodal models is further enhancing this advancement. However, there is a lack of standardized and user-friendly tools for research and development, as well as experimental platforms on live websites. To address this challenge, we present WebOlympus, an open platform for web agents operating on live websites. WebOlympus offers a Chrome extension-based UI, enabling users without programming experience to easily utilize the platform. It allows users to run web agents with various designs using only a few lines of code or simple clicks on the Chrome extension. To ensure the trustworthiness of web agents, a safety monitor module that prevents harmful actions through human supervision or model-based control is incorporated. WebOlympus supports diverse applications, including annotation interfaces for web agent trajectories and data crawling. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-demo.42 | null | 1 | general safety, LLM alignment | null | WalledEval: A Comprehensive Safety Evaluation Toolkit for Large Language Models | WalledEval is a comprehensive AI safety testing toolkit designed to evaluate large language models (LLMs). It accommodates a diverse range of models, including both open-weight and API-based ones, and features over 35 safety benchmarks covering areas such as multilingual safety, exaggerated safety, and prompt injections. The framework supports both LLM and judge benchmarking, and incorporates custom mutators to test safety against various text-style mutations such as future tense and paraphrasing. Additionally, WalledEval introduces WalledGuard, a new, small and performant content moderation tool, and SGXSTest, a benchmark for assessing exaggerated safety in cultural contexts. We make WalledEval publicly available at https://github.com/walledai/walledeval with a demonstration video at https://youtu.be/50Zy97kj1MA. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-industry.16 | null | 0 | toxicity, bias | null | Debiasing Text Safety Classifiers through a Fairness-Aware Ensemble | Increasing use of large language models (LLMs) demand performant guardrails to ensure the safety of inputs and outputs of LLMs. When these safeguards are trained on imbalanced data, they can learn the societal biases. We present a light-weight, post-processing method for mitigating counterfactual fairness in closed-source text safety classifiers. Our approach involves building an ensemble that not only outperforms the input classifiers and policy-aligns them, but also acts as a debiasing regularizer. We introduce two threshold-agnostic metrics to assess the counterfactual fairness of a model, and demonstrate how combining these metrics with Fair Data Reweighting (FDW) helps mitigate biases. We create an expanded Open AI dataset, and a new templated LLM-generated dataset based on user-prompts, both of which are counterfactually balanced across identity groups and cover four key areas of safety; we will work towards publicly releasing these datasets. Our results show that our approach improves counterfactual fairness with minimal impact on model performance. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-industry.57 | x | 0 | privacy | null | ULMR: Unlearning Large Language Models via Negative Response and Model Parameter Average | In recent years, large language models (LLMs) have attracted significant interest from the research community due to their broad applicability in many language-oriented tasks, and are now widely used in numerous areas of production and daily life. One source of the powerful capabilities of LLMs is the massive scale of their pre-training dataset. However, these pre-training datasets contain many outdated, harmful, and personally sensitive information, which inevitably becomes memorized by LLM during the pre-training process. Eliminating this undesirable data is crucial for ensuring the model’s safety and enhancing the user experience. However, the cost of extensively cleaning the pre-training dataset and retraining the model from scratch is very high. In this work, we propose ULMR , a unlearning framework for LLMs , which first uses carefully designed prompts to rewrite the instructions in the specified dataset, and generate corresponding negative responses. Subsequently, to ensure that the model does not excessively deviate post-training, we perform model parameter averaging to preserve the performance of the original LLM. We conducted experiments on two public datasets, TOFU and RWKU, demonstrating that our method can effectively forget specified information while retaining the capabilities of the original LLM. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-industry.76 | x | 0 | others | null | Survival of the Safest: Towards Secure Prompt Optimization through Interleaved Multi-Objective Evolution | Large language models (LLMs) have demonstrated remarkable capabilities; however, the optimization of their prompts has historically prioritized performance metrics at the expense of crucial safety and security considerations. To overcome this shortcoming, we introduce “Survival of the Safest” (), an innovative multi-objective prompt optimization framework that enhances both performance and security in LLMs simultaneously. utilizes an interleaved multi-objective evolution strategy, integrating semantic, feedback, and crossover mutations to effectively traverse the prompt landscape. Differing from the computationally demanding Pareto front methods, provides a scalable solution that expedites optimization in complex, high-dimensional discrete search spaces while keeping computational demands low. Our approach accommodates flexible weighting of objectives and generates a pool of optimized candidates, empowering users to select prompts that optimally meet their specific performance and security needs. Experimental evaluations across diverse benchmark datasets affirm ‘s efficacy in delivering high performance and notably enhancing safety and security compared to single-objective methods. This advancement marks a significant stride towards the deployment of LLM systems that are both high-performing and secure across varied industrial applications |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-industry.84 | x | 0 | general safety, LLM alignment | null | Athena: Safe Autonomous Agents with Verbal Contrastive Learning | Due to emergent capabilities, large language models (LLMs) have been utilized as language-based agents to perform a variety of tasks and make decisions with an increasing degree of autonomy. These autonomous agents can understand high-level instructions, interact with their environments, and execute complex tasks using a selection of tools available to them. As the capabilities of the agents expand, ensuring their safety and trustworthiness becomes more imperative. In this study, we introduce the Athena framework which leverages the concept of verbal contrastive learning where past safe and unsafe trajectories are used as in-context (contrastive) examples to guide the agent towards safety while fulfilling a given task. The framework also incorporates a critiquing mechanism to guide the agent to prevent risky actions at every step. Furthermore, due to the lack of existing benchmarks on the safety reasoning ability of LLM-based agents, we curate a set of 80 toolkits across 8 categories with 180 scenarios to provide a safety evaluation benchmark. Our experimental evaluation, with both closed- and open-source LLMs, indicates verbal contrastive learning and interaction-level critiquing improve the safety rate significantly. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-industry.99 | null | 1 | general safety, LLM alignment | English and Korean | SLM as Guardian: Pioneering AI Safety with Small Language Model | Most prior safety research of large language models (LLMs) has focused on enhancing the alignment of LLMs to better suit the safety requirements of their use cases. However, internalizing such safeguard features into larger models brought challenges of higher training cost and unintended degradation of helpfulness. In this paper, we leverage a smaller LLM for both harmful query detection and safeguard response generation. We introduce our safety requirements and the taxonomy of harmfulness categories, and then propose a multi-task learning mechanism fusing the two tasks into a single model. We demonstrate the effectiveness of our approach, providing on par or surpassing harmful query detection and safeguard response performance compared to the publicly available LLMs. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-industry.115 | x | 0 | others | null | Aegis:An Advanced LLM-Based Multi-Agent for Intelligent Functional Safety Engineering | Functional safety is a critical aspect of automotive engineering, encompassing all phases of a vehicle’s lifecycle, including design, development, production, operation, and decommissioning. This domain involves highly knowledge-intensive tasks. This paper introduces Aegis: An Advanced LLM-Based Multi-Agent for Intelligent Functional Safety Engineering. Aegis is specifically designed to support complex functional safety tasks within the automotive sector. It is tailored to perform Hazard Analysis and Risk Assessment (HARA), document Functional Safety Requirements (FSR), and plan test cases for Automatic Emergency Braking (AEB) systems. The most advanced version, Aegis-Max, leverages Retrieval-Augmented Generation (RAG) and reflective mechanisms to enhance its capability in managing complex, knowledge-intensive tasks. Additionally, targeted prompt refinement by professional functional safety practitioners can significantly optimize Aegis’s performance in the functional safety domain. This paper demonstrates the potential of Aegis to improve the efficiency and effectiveness of functional safety processes in automotive engineering. |
2024.emnlp.xml | Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing | 2024 | https://aclanthology.org/2024.emnlp-industry.119 | null | 0 | jailbreaking attacks | null | Course-Correction: Safety Alignment Using Synthetic Preferences | The risk of harmful contents generated by large language models (LLMs) becomes a critical concern. This paper systematically evaluates and enhances LLMs’ capability to perform course-correction, , the model can steer away from generating harmful content autonomously. First, we introduce the C^2-Eval benchmark for quantitative assessment and analyze 10 popular LLMs, revealing varying proficiency of current safety-tuned LLMs in course-correction.To improve, we propose fine-tuning LLMs with preference learning, emphasizing the preference for timely course-correction. Using an automated pipeline, we create C^2-Syn, a synthetic C^2-Syn with 750K pairwise preferences, to teach models the concept of timely course-correction through data-driven learning.Experiments on Llama2-Chat 7B and Qwen2 7B show that our method effectively enhances course-correction skills without affecting general performance. Additionally, it effectively improves LLMs’ safety, particularly in resisting jailbreak attacks. |
2024.nejlt.xml | Northern European Journal of Language Technology, Volume 10 | 2024 | https://aclanthology.org/2024.nejlt-1.5 | x | 0 | others | null | On Using Self-Report Studies to Analyze Language Models | We are at a curious point in time where our ability to build language models (LMs) has outpaced our ability to analyze them. We do not really know how to reliably determine their capabilities, biases, dangers, knowledge, and so on. The benchmarks we have are often overly specific, do not generalize well, and are susceptible to data leakage. Recently, I have noticed a trend of using self-report studies, such as various polls and questionnaires originally designed for humans, to analyze the properties of LMs. I think that this approach can easily lead to false results, which can be quite dangerous considering the current discussions on AI safety, governance, and regulation. To illustrate my point, I will delve deeper into several papers that employ self-report methodologies and I will try to highlight some of their weaknesses. |
2024.yrrsds.xml | Proceedings of the 20th Workshop of Young Researchers' Roundtable on Spoken Dialogue Systems | 2024 | https://aclanthology.org/2024.yrrsds-1.24 | x | 0 | others | null | Innovative Approaches to Enhancing Safety and Ethical AI Interactions in Digital Environments | Ensuring safe online environments is a formidable challenge, but nonetheless an important one as people are now chronically online. The increasing online presence of people paired with the prevalence of harmful content such as toxicity, hate speech, misinformation and disinformation across various social media platforms and within different video calls for stronger detection and prevention methods. My research interests primarily lie in applied natural language processing for social good. Previously, I focused on measuring partisan polarization on social media during the COVID-19 pandemic and its societal impacts. Currently, at Ubisoft La Forge, I am dedicated to enhancing player safety within in-game chat systems by developing methods to detect toxicity, evaluating the biases in these detection systems, and assessing the current ecological state of online interactions. Additionally, I am engaged in simulating social media environments using LLMs to ethically test detection methods, evaluate the effectiveness of current mitigation strategies, and potentially introduce new, successful strategies. My suggested topics for discussion: 1. Understanding and mitigating social harms through high fidelity simulated social media environments 2. Enhancing safety in online environments such as within in-game chats (text and speech) 3. Personification of LLM agents 4. Ethically simulating social media sandbox environments at scale with LLM agents 5. Re-balancing the playing field between good and bad actors: Strategies for countering societal-scale manipulation. |
2024.nlp4pi.xml | Proceedings of the Third Workshop on NLP for Positive Impact | 2024 | https://aclanthology.org/2024.nlp4pi-1.7 | null | 0 | others | null | PG-Story: Taxonomy, Dataset, and Evaluation for Ensuring Child-Safe Content for Story Generation | Creating children’s stories through text generation is a creative task that requires stories to be both entertaining and suitable for young audiences. However, since current story generation systems often rely on pre-trained language models fine-tuned with limited story data, they may not always prioritize child-friendliness. This can lead to the unintended generation of stories containing problematic elements such as violence, profanity, and biases. Regrettably, despite the significance of these concerns, there is a lack of clear guidelines and benchmark datasets for ensuring content safety for children. In this paper, we introduce a taxonomy specifically tailored to assess content safety in text, with a strong emphasis on children’s well-being. We present PG-Story, a dataset that includes detailed annotations for both sentence-level and discourse-level safety. We demonstrate the potential of identifying unsafe content through self-diagnosis and employing controllable generation techniques during the decoding phase to minimize unsafe elements in generated stories. |
2024.nlp4pi.xml | Proceedings of the Third Workshop on NLP for Positive Impact | 2024 | https://aclanthology.org/2024.nlp4pi-1.30 | x | 0 | general safety, LLM alignment | null | Improving Industrial Safety by Auto-Generating Case-specific Preventive Recommendations | In this paper, we propose a novel application to improve industrial safety by generating preventive recommendations using LLMs. Using a dataset of 275 incidents representing 11 different incident types sampled from real-life OSHA incidents, we compare three different LLMs to evaluate the quality of preventive recommendations generated by them. We also show that LLMs are not a panacea for the preventive recommendation generation task. They have limitations and can produce responses that are incorrect or irrelevant. We found that about 65% of the output from Vicuna model was not acceptable at all at the basic readability and other sanity checks level. Mistral and Phi_3 are better than Vicuna, but not all of their recommendations are of similar quality. We find that for a given safety incident case, the generated recommendations can be categorized as specific, generic, or irrelevant. This helps us to better quantify and compare the performance of the models. This paper is among the initial and novel work for the preventive recommendation generation problem. We believe it will pave way for use of NLP to positively impact the industrial safety. |
2022.lrec.xml | Proceedings of the Thirteenth Language Resources and Evaluation Conference | 2022 | https://aclanthology.org/2022.lrec-1.668 | null | 0 | others | null | TWEET-FID: An Annotated Dataset for Multiple Foodborne Illness Detection Tasks | Foodborne illness is a serious but preventable public health problem – with delays in detecting the associated outbreaks resulting in productivity loss, expensive recalls, public safety hazards, and even loss of life. While social media is a promising source for identifying unreported foodborne illnesses, there is a dearth of labeled datasets for developing effective outbreak detection models. To accelerate the development of machine learning-based models for foodborne outbreak detection, we thus present TWEET-FID (TWEET-Foodborne Illness Detection), the first publicly available annotated dataset for multiple foodborne illness incident detection tasks. TWEET-FID collected from Twitter is annotated with three facets: tweet class, entity type, and slot type, with labels produced by experts as well as by crowdsource workers. We introduce several domain tasks leveraging these three facets: text relevance classification (TRC), entity mention detection (EMD), and slot filling (SF). We describe the end-to-end methodology for dataset design, creation, and labeling for supporting model development for these tasks. A comprehensive set of results for these tasks leveraging state-of-the-art single-and multi-task deep learning methods on the TWEET-FID dataset are provided. This dataset opens opportunities for future research in foodborne outbreak detection. |
2022.lrec.xml | Proceedings of the Thirteenth Language Resources and Evaluation Conference | 2022 | https://aclanthology.org/2022.lrec-1.673 | x | 0 | others | null | Knowledge Graph - Deep Learning: A Case Study in Question Answering in Aviation Safety Domain | In the commercial aviation domain, there are a large number of documents, like accident reports of NTSB and ASRS, and regulatory directives ADs. There is a need for a system to efficiently access these diverse repositories to serve the demands of the aviation industry, such as maintenance, compliance, and safety. In this paper, we propose a Knowledge Graph (KG) guided Deep Learning (DL) based Question Answering (QA) system to cater to these requirements. We construct a KG from aircraft accident reports and contribute this resource to the community of researchers. The efficacy of this resource is tested and proved by the proposed QA system. Questions in Natural Language are converted into SPARQL (the interface language of the RDF graph database) queries and are answered from the KG. On the DL side, we examine two different QA models, BERT-QA and GPT3-QA, covering the two paradigms of answer formulation in QA. We evaluate our system on a set of handcrafted queries curated from the accident reports. Our hybrid KG + DL QA system, KGQA + BERT-QA, achieves 7% and 40.3% increase in accuracy over KGQA and BERT-QA systems respectively. Similarly, the other combined system, KGQA + GPT3-QA, achieves 29.3% and 9.3% increase in accuracy over KGQA and GPT3-QA systems respectively. Thus, we infer that the combination of KG and DL is better than either KG or DL individually for QA, at least in our chosen domain. |
2022.lrec.xml | Proceedings of the Thirteenth Language Resources and Evaluation Conference | 2022 | https://aclanthology.org/2022.lrec-1.731 | null | 1 | others | English, Cantonese | CI-AVSR: A Cantonese Audio-Visual Speech Datasetfor In-car Command Recognition | With the rise of deep learning and intelligent vehicles, the smart assistant has become an essential in-car component to facilitate driving and provide extra functionalities. In-car smart assistants should be able to process general as well as car-related commands and perform corresponding actions, which eases driving and improves safety. However, there is a data scarcity issue for low resource languages, hindering the development of research and applications. In this paper, we introduce a new dataset, Cantonese In-car Audio-Visual Speech Recognition (CI-AVSR), for in-car command recognition in the Cantonese language with both video and audio data. It consists of 4,984 samples (8.3 hours) of 200 in-car commands recorded by 30 native Cantonese speakers. Furthermore, we augment our dataset using common in-car background noises to simulate real environments, producing a dataset 10 times larger than the collected one. We provide detailed statistics of both the clean and the augmented versions of our dataset. Moreover, we implement two multimodal baselines to demonstrate the validity of CI-AVSR. Experiment results show that leveraging the visual signal improves the overall performance of the model. Although our best model can achieve a considerable quality on the clean test set, the speech recognition quality on the noisy data is still inferior and remains an extremely challenging task for real in-car speech recognition systems. The dataset and code will be released at https://github.com/HLTCHKUST/CI-AVSR. |
2022.acl.xml | Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) | 2022 | https://aclanthology.org/2022.acl-long.284 | null | 0 | general safety, LLM alignment | null | SafetyKit: First Aid for Measuring Safety in Open-domain Conversational Systems | The social impact of natural language processing and its applications has received increasing attention. In this position paper, we focus on the problem of safety for end-to-end conversational AI. We survey the problem landscape therein, introducing a taxonomy of three observed phenomena: the Instigator, Yea-Sayer, and Impostor effects. We then empirically assess the extent to which current tools can measure these effects and current systems display them. We release these tools as part of a “first aid kit” (SafetyKit) to quickly assess apparent safety concerns. Our results show that, while current tools are able to provide an estimate of the relative safety of systems in various settings, they still have several shortcomings. We suggest several future directions and discuss ethical considerations. |
2022.acl.xml | Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) | 2022 | https://aclanthology.org/2022.acl-long.394 | x | 0 | others | null | Human Evaluation and Correlation with Automatic Metrics in Consultation Note Generation | In recent years, machine learning models have rapidly become better at generating clinical consultation notes; yet, there is little work on how to properly evaluate the generated consultation notes to understand the impact they may have on both the clinician using them and the patient’s clinical safety. To address this we present an extensive human evaluation study of consultation notes where 5 clinicians (i) listen to 57 mock consultations, (ii) write their own notes, (iii) post-edit a number of automatically generated notes, and (iv) extract all the errors, both quantitative and qualitative. We then carry out a correlation study with 18 automatic quality metrics and the human judgements. We find that a simple, character-based Levenshtein distance metric performs on par if not better than common model-based metrics like BertScore. All our findings and annotations are open-sourced. |
2022.acl.xml | Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) | 2022 | https://aclanthology.org/2022.acl-long.447 | null | 0 | general safety, LLM alignment | null | SaFeRDialogues: Taking Feedback Gracefully after Conversational Safety Failures | Current open-domain conversational models can easily be made to talk in inadequate ways. Online learning from conversational feedback given by the conversation partner is a promising avenue for a model to improve and adapt, so as to generate fewer of these safety failures. However, current state-of-the-art models tend to react to feedback with defensive or oblivious responses. This makes for an unpleasant experience and may discourage conversation partners from giving feedback in the future. This work proposes SaFeRDialogues, a task and dataset of graceful responses to conversational feedback about safety failures. We collect a dataset of 8k dialogues demonstrating safety failures, feedback signaling them, and a response acknowledging the feedback. We show how fine-tuning on this dataset results in conversations that human raters deem considerably more likely to lead to a civil conversation, without sacrificing engagingness or general conversational ability. |
2022.acl.xml | Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) | 2022 | https://aclanthology.org/2022.acl-long.538 | x | 0 | jailbreaking attacks | null | “That Is a Suspicious Reaction!”: Interpreting Logits Variation to Detect NLP Adversarial Attacks | Adversarial attacks are a major challenge faced by current machine learning research. These purposely crafted inputs fool even the most advanced models, precluding their deployment in safety-critical applications. Extensive research in computer vision has been carried to develop reliable defense strategies. However, the same issue remains less explored in natural language processing. Our work presents a model-agnostic detector of adversarial text examples. The approach identifies patterns in the logits of the target classifier when perturbing the input text. The proposed detector improves the current state-of-the-art performance in recognizing adversarial inputs and exhibits strong generalization capabilities across different NLP models, datasets, and word-level attacks. |
2022.naacl.xml | Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies | 2022 | https://aclanthology.org/2022.naacl-main.155 | null | 1 | general safety, LLM alignment | Korean | Building a Role Specified Open-Domain Dialogue System Leveraging Large-Scale Language Models | Recent open-domain dialogue models have brought numerous breakthroughs. However, building a chat system is not scalable since it often requires a considerable volume of human-human dialogue data, especially when enforcing features such as persona, style, or safety. In this work, we study the challenge of imposing roles on open-domain dialogue systems, with the goal of making the systems maintain consistent roles while conversing naturally with humans. To accomplish this, the system must satisfy a role specification that includes certain conditions on the stated features as well as a system policy on whether or not certain types of utterances are allowed. For this, we propose an efficient data collection framework leveraging in-context few-shot learning of large-scale language models for building role-satisfying dialogue dataset from scratch. We then compare various architectures for open-domain dialogue systems in terms of meeting role specifications while maintaining conversational abilities. Automatic and human evaluations show that our models return few out-of-bounds utterances, keeping competitive performance on general metrics. We release a Korean dialogue dataset we built for further research. |
2022.naacl.xml | Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies | 2022 | https://aclanthology.org/2022.naacl-main.363 | null | 0 | toxicity, bias | null | Mitigating Toxic Degeneration with Empathetic Data: Exploring the Relationship Between Toxicity and Empathy | Large pre-trained neural language models have supported the effectiveness of many NLP tasks, yet are still prone to generating toxic language hindering the safety of their use. Using empathetic data, we improve over recent work on controllable text generation that aims to reduce the toxicity of generated text. We find we are able to dramatically reduce the size of fine-tuning data to 7.5-30k samples while at the same time making significant improvements over state-of-the-art toxicity mitigation of up to 3.4% absolute reduction (26% relative) from the original work on 2.3m samples, by strategically sampling data based on empathy scores. We observe that the degree of improvements is subject to specific communication components of empathy. In particular, the more cognitive components of empathy significantly beat the original dataset in almost all experiments, while emotional empathy was tied to less improvement and even underperforming random samples of the original data. This is a particularly implicative insight for NLP work concerning empathy as until recently the research and resources built for it have exclusively considered empathy as an emotional concept. |
2022.tacl.xml | Transactions of the Association for Computational Linguistics, Volume 10 | 2022 | https://aclanthology.org/2022.tacl-1.39 | x | 0 | others | null | Uncertainty Estimation and Reduction of Pre-trained Models for Text Regression | State-of-the-art classification and regression models are often not well calibrated, and cannot reliably provide uncertainty estimates, limiting their utility in safety-critical applications such as clinical decision-making. While recent work has focused on calibration of classifiers, there is almost no work in NLP on calibration in a regression setting. In this paper, we quantify the calibration of pre- trained language models for text regression, both intrinsically and extrinsically. We further apply uncertainty estimates to augment training data in low-resource domains. Our experiments on three regression tasks in both self-training and active-learning settings show that uncertainty estimation can be used to increase overall performance and enhance model generalization. |
2022.findings.xml | Findings of the Association for Computational Linguistics: ACL 2022 | 2022 | https://aclanthology.org/2022.findings-acl.185 | null | null | others | null | Systematicity, Compositionality and Transitivity of Deep NLP Models: a Metamorphic Testing Perspective | Metamorphic testing has recently been used to check the safety of neural NLP models. Its main advantage is that it does not rely on a ground truth to generate test cases. However, existing studies are mostly concerned with robustness-like metamorphic relations, limiting the scope of linguistic properties they can test. We propose three new classes of metamorphic relations, which address the properties of systematicity, compositionality and transitivity. Unlike robustness, our relations are defined over multiple source inputs, thus increasing the number of test cases that we can produce by a polynomial factor. With them, we test the internal consistency of state-of-the-art NLP models, and show that they do not always behave according to their expected linguistic properties. Lastly, we introduce a novel graphical notation that efficiently summarises the inner structure of metamorphic relations. |
2022.findings.xml | Findings of the Association for Computational Linguistics: ACL 2022 | 2022 | https://aclanthology.org/2022.findings-acl.308 | null | 0 | general safety, LLM alignment | null | On the Safety of Conversational Models: Taxonomy, Dataset, and Benchmark | Dialogue safety problems severely limit the real-world deployment of neural conversational models and have attracted great research interests recently. However, dialogue safety problems remain under-defined and the corresponding dataset is scarce. We propose a taxonomy for dialogue safety specifically designed to capture unsafe behaviors in human-bot dialogue settings, with focuses on context-sensitive unsafety, which is under-explored in prior works. To spur research in this direction, we compile DiaSafety, a dataset with rich context-sensitive unsafe examples. Experiments show that existing safety guarding tools fail severely on our dataset. As a remedy, we train a dialogue safety classifier to provide a strong baseline for context-sensitive dialogue unsafety detection. With our classifier, we perform safety evaluations on popular conversational models and show that existing dialogue systems still exhibit concerning context-sensitive safety problems. |
2022.findings.xml | Findings of the Association for Computational Linguistics: ACL 2022 | 2022 | https://aclanthology.org/2022.findings-emnlp.211 | x | 0 | toxicity, bias | null | Mitigating Covertly Unsafe Text within Natural Language Systems | An increasingly prevalent problem for intelligent technologies is text safety, as uncontrolled systems may generate recommendations to their users that lead to injury or life-threatening consequences. However, the degree of explicitness of a generated statement that can cause physical harm varies. In this paper, we distinguish types of text that can lead to physical harm and establish one particularly underexplored category: covertly unsafe text. Then, we further break down this category with respect to the system’s information and discuss solutions to mitigate the generation of text in each of these subcategories. Ultimately, our work defines the problem of covertly unsafe language that causes physical harm and argues that this subtle yet dangerous issue needs to be prioritized by stakeholders and regulators. We highlight mitigation strategies to inspire future researchers to tackle this challenging problem and help improve safety within smart systems. |
2022.findings.xml | Findings of the Association for Computational Linguistics: ACL 2022 | 2022 | https://aclanthology.org/2022.findings-emnlp.262 | null | 1 | toxicity, bias | Chinese | Towards Identifying Social Bias in Dialog Systems: Framework, Dataset, and Benchmark | Among all the safety concerns that hinder the deployment of open-domain dialog systems (e.g., offensive languages, biases, and toxic behaviors), social bias presents an insidious challenge. Addressing this challenge requires rigorous analyses and normative reasoning. In this paper, we focus our investigation on social bias measurement to facilitate the development of unbiased dialog systems. We first propose a novel Dial-Bias Framework for analyzing the social bias in conversations using a holistic method beyond bias lexicons or dichotomous annotations. Leveraging the proposed framework, we further introduce the CDial-Bias Dataset which is, to the best of our knowledge, the first annotated Chinese social bias dialog dataset. We also establish a fine-grained dialog bias measurement benchmark and conduct in-depth ablation studies to shed light on the utility of the detailed annotations in the proposed dataset. Finally, we evaluate representative Chinese generative models with our classifiers to unveil the presence of social bias in these systems. |
2022.findings.xml | Findings of the Association for Computational Linguistics: ACL 2022 | 2022 | https://aclanthology.org/2022.findings-emnlp.270 | x | 0 | toxicity, bias | null | Constructing Highly Inductive Contexts for Dialogue Safety through Controllable Reverse Generation | Large pretrained language models can easily produce toxic or biased content, which is prohibitive for practical use. In order to detect such toxic generations, existing methods rely on templates, real-world data extraction, crowdsourcing workers or automatic generation to construct adversarial contexts that are likely to induce toxic generations. However, what type of context is more likely to induce unsafe responses is still under-explored. In this paper, we identify that context toxicity and context category (e.g., profanity, insult, drugs, etc.) are two important factors to cause safety issues in response generation. Hence, we propose a method called reverse generation to construct adversarial contexts conditioned on a given response, with the flexibility to control category, toxicity level and inductivity of the generated contexts. Via reverse generation, we augment the existing BAD dataset and construct a new dataset BAD+ which contains more than 120K diverse and highly inductive contexts in 12 categories. We test three popular pretrained dialogue models (Blender, DialoGPT and Plato2) and find that BAD+ can largely expose their safety problems. Furthermore, we show that BAD+ can greatly enhance the safety of generation, and we reveal the key factors of safety improvement. Our code and dataset is available at https://github.com/thu-coai/Reverse_Generation. |
2022.findings.xml | Findings of the Association for Computational Linguistics: ACL 2022 | 2022 | https://aclanthology.org/2022.findings-emnlp.538 | null | null | others | null | Uncertainty Quantification with Pre-trained Language Models: A Large-Scale Empirical Analysis | Pre-trained language models (PLMs) have gained increasing popularity due to their compelling prediction performance in diverse natural language processing (NLP) tasks. When formulating a PLM-based prediction pipeline for NLP tasks, it is also crucial for the pipeline to minimize the calibration error, especially in safety-critical applications. That is, the pipeline should reliably indicate when we can trust its predictions. In particular, there are various considerations behind the pipeline: (1) the choice and (2) the size of PLM, (3) the choice of uncertainty quantifier, (4) the choice of fine-tuning loss, and many more. Although prior work has looked into some of these considerations, they usually draw conclusions based on a limited scope of empirical studies. There still lacks a holistic analysis on how to compose a well-calibrated PLM-based prediction pipeline. To fill this void, we compare a wide range of popular options for each consideration based on three prevalent NLP classification tasks and the setting of domain shift. In response, we recommend the following: (1) use ELECTRA for PLM encoding, (2) use larger PLMs if possible, (3) use Temp Scaling as the uncertainty quantifier, and (4) use Focal Loss for fine-tuning. |
2022.icon.xml | Proceedings of the 19th International Conference on Natural Language Processing (ICON) | 2022 | https://aclanthology.org/2022.icon-main.26 | null | null | others | null | There is No Big Brother or Small Brother:Knowledge Infusion in Language Models for Link Prediction and Question Answering | The integration of knowledge graphs with deep learning is thriving in improving the performance of various natural language processing (NLP) tasks. In this paper, we focus on knowledge-infused link prediction and question answering using language models, T5, and BLOOM across three domains:Aviation, Movie, and Web. In this context, we infuse knowledge in large and small language models and study their performance, and find the performance to be similar. For the link prediction task on the Aviation Knowledge Graph, we obtain a 0.2 hits@1 score using T5-small, T5-base, T5-large, and BLOOM. Using template-based scripts, we create a set of 1 million synthetic factoid QA pairs in the aviation domain from National Transportation Safety Board (NTSB) reports. On our curated QA pairs, the three models of T5 achieve a 0.7 hits@1 score. We validate our findings with the paired student t test and Cohen’s kappa scores. For link prediction on Aviation Knowledge Graph using T5-small and T5-large, we obtain a Cohen’s kappa score of 0.76, showing substantial agreement between the models. Thus, we infer that small language models perform similar to large language models with the infusion of knowledge. |
2022.emnlp.xml | Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing | 2022 | https://aclanthology.org/2022.emnlp-main.154 | null | 0 | general safety, LLM alignment | null | SafeText: A Benchmark for Exploring Physical Safety in Language Models | Understanding what constitutes safe text is an important issue in natural language processing and can often prevent the deployment of models deemed harmful and unsafe. One such type of safety that has been scarcely studied is commonsense physical safety, i.e. text that is not explicitly violent and requires additional commonsense knowledge to comprehend that it leads to physical harm. We create the first benchmark dataset, SafeText, comprising real-life scenarios with paired safe and physically unsafe pieces of advice. We utilize SafeText to empirically study commonsense physical safety across various models designed for text generation and commonsense reasoning tasks. We find that state-of-the-art large language models are susceptible to the generation of unsafe text and have difficulty rejecting unsafe advice. As a result, we argue for further studies of safety and the assessment of commonsense physical safety in models before release. |
2022.emnlp.xml | Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing | 2022 | https://aclanthology.org/2022.emnlp-main.267 | null | 0 | general safety, LLM alignment | null | ProsocialDialog: A Prosocial Backbone for Conversational Agents | Most existing dialogue systems fail to respond properly to potentially unsafe user utterances by either ignoring or passively agreeing with them. To address this issue, we introduce ProsocialDialog, the first large-scale multi-turn dialogue dataset to teach conversational agents to respond to problematic content following social norms. Covering diverse unethical, problematic, biased, and toxic situations, ProsocialDialog contains responses that encourage prosocial behavior, grounded in commonsense social rules (i.e., rules-of-thumb, RoTs). Created via a human-AI collaborative framework, ProsocialDialog consists of 58K dialogues, with 331K utterances, 160K unique RoTs, and 497K dialogue safety labels accompanied by free-form rationales.With this dataset, we introduce a dialogue safety detection module, Canary, capable of generating RoTs given conversational context, and a socially-informed dialogue agent, Prost. Empirical results show that Prost generates more socially acceptable dialogues compared to other state-of-the-art language and dialogue models in both in-domain and out-of-domain settings. Additionally, Canary effectively guides conversational agents and off-the-shelf language models to generate significantly more prosocial responses. Our work highlights the promise and importance of creating and steering conversational AI to be socially responsible. |
2022.emnlp.xml | Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing | 2022 | https://aclanthology.org/2022.emnlp-main.376 | null | null | others | null | PHEE: A Dataset for Pharmacovigilance Event Extraction from Text | The primary goal of drug safety researchers and regulators is to promptly identify adverse drug reactions. Doing so may in turn prevent or reduce the harm to patients and ultimately improve public health. Evaluating and monitoring drug safety (i.e., pharmacovigilance) involves analyzing an ever growing collection of spontaneous reports from health professionals, physicians, and pharmacists, and information voluntarily submitted by patients. In this scenario, facilitating analysis of such reports via automation has the potential to rapidly identify safety signals. Unfortunately, public resources for developing natural language models for this task are scant. We present PHEE, a novel dataset for pharmacovigilance comprising over 5000 annotated events from medical case reports and biomedical literature, making it the largest such public dataset to date. We describe the hierarchical event schema designed to provide coarse and fine-grained information about patients’ demographics, treatments and (side) effects. Along with the discussion of the dataset, we present a thorough experimental evaluation of current state-of-the-art approaches for biomedical event extraction, point out their limitations, and highlight open challenges to foster future research in this area. |
2022.semeval.xml | Proceedings of the 16th International Workshop on Semantic Evaluation (SemEval-2022) | 2022 | https://aclanthology.org/2022.semeval-1.75 | null | null | toxicity, bias | null | Transformers at SemEval-2022 Task 5: A Feature Extraction based Approach for Misogynous Meme Detection | Social media is an idea created to make theworld smaller and more connected. Recently,it has become a hub of fake news and sexistmemes that target women. Social Media shouldensure proper women’s safety and equality. Filteringsuch information from social media is ofparamount importance to achieving this goal. In this paper, we describe the system developedby our team for SemEval-2022 Task 5: MultimediaAutomatic Misogyny Identification. Wepropose a multimodal training methodologythat achieves good performance on both thesubtasks, ranking 4th for Subtask A (0.718macro F1-score) and 9th for Subtask B (0.695macro F1-score) while exceeding the baselineresults by good margins. |
2022.legal.xml | Proceedings of the Workshop on Ethical and Legal Issues in Human Language Technologies and Multilingual De-Identification of Sensitive Data In Language Resources within the 13th Language Resources and Evaluation Conference | 2022 | https://aclanthology.org/2022.legal-1.2 | null | null | others | null | Sentiment Analysis and Topic Modeling for Public Perceptions of Air Travel: COVID Issues and Policy Amendments | Among many industries, air travel is impacted by the COVID pandemic. Airlines and airports rely on public sector information to enforce guidelines for ensuring health and safety of travelers. Such guidelines can be policy amendments or laws during the pandemic. In response to the inception of COVID preventive policies, travelers have exercised freedom of expression via the avenue of online reviews. This avenue facilitates voicing public concern while anonymizing / concealing user identity as needed. It is important to assess opinions on policy amendments to ensure transparency and openness, while also preserving confidentiality and ethics. Hence, this study leverages data science to analyze, with identity protection, the online reviews of airlines and airports since 2017, considering impacts of COVID issues and relevant policy amendments since 2020. Supervised learning with VADER sentiment analysis is deployed to predict changes in opinion from 2017 to date. Unsupervised learning with LDA topic modeling is employed to discover air travelers’ major areas of concern before and after the pandemic. This study reveals that COVID policies have worsened public perceptions of air travel and aroused notable new concerns, affecting economics, environment and health. |
2022.clpsych.xml | Proceedings of the Eighth Workshop on Computational Linguistics and Clinical Psychology | 2022 | https://aclanthology.org/2022.clpsych-1.2 | x | 0 | general safety, LLM alignment | null | The ethical role of computational linguistics in digital psychological formulation and suicide prevention. | Formulation is central to clinical practice. Formulation has a factor weighing, pattern recognition and explanatory hypothesis modelling focus. Formulation attempts to make sense of why a person presents in a certain state at a certain time and context, and how that state may be best managed to enhance mental health, safety and optimal change. Inherent to the clinical need for formulation is an appreciation of the complexities, uncertainty and limits of applying theoretical concepts and symptom, diagnostic and risk categories to human experience; or attaching meaning or weight to any particular factor in an individual?s history or mental state without considering the broader biopsychosocial and cultural context. With specific reference to suicide prevention, this paper considers the need and potential for the computer linguistic community to be both cognisant of and ethically contribute to the clinical formulation process. |
2022.computel.xml | Proceedings of the Fifth Workshop on the Use of Computational Methods in the Study of Endangered Languages | 2022 | https://aclanthology.org/2022.computel-1.13 | null | null | others | null | Challenges and Perspectives for Innu-Aimun within Indigenous Language Technologies | Innu-Aimun is an Algonquian language spoken in Eastern Canada. It is the language of the Innu, an Indigenous people that now lives for the most part in a dozen communities across Quebec and Labrador. Although it is alive, Innu-Aimun sees important preservation and revitalization challenges and issues. The state of its technology is still nascent, with very few existing applications. This paper proposes a first survey of the available linguistic resources and existing technology for Innu-Aimun. Considering the existing linguistic and textual resources, we argue that developing language technology is feasible and propose first steps towards NLP applications like machine translation. The goal of developing such technologies is first and foremost to help efforts in improving language transmission and cultural safety and preservation for Innu-Aimun speakers, as those are considered urgent and vital issues. Finally, we discuss the importance of close collaboration and consultation with the Innu community in order to ensure that language technologies are developed respectfully and in accordance with that goal. |
2022.wmt.xml | Proceedings of the Seventh Conference on Machine Translation (WMT) | 2022 | https://aclanthology.org/2022.wmt-1.43 | null | null | others | null | Robust MT Evaluation with Sentence-level Multilingual Augmentation | Automatic translations with critical errors may lead to misinterpretations and pose several risks for the user. As such, it is important that Machine Translation (MT) Evaluation systems are robust to these errors in order to increase the reliability and safety of Machine Translation systems. Here we introduce SMAUG a novel Sentence-level Multilingual AUGmentation approach for generating translations with critical errors and apply this approach to create a test set to evaluate the robustness of MT metrics to these errors. We show that current State-of-the-Art metrics are improving their capability to distinguish translations with and without critical errors and to penalize the first accordingly. We also show that metrics tend to struggle with errors related to named entities and numbers and that there is a high variance in the robustness of current methods to translations with critical errors. |
2022.coling.xml | Proceedings of the 29th International Conference on Computational Linguistics | 2022 | https://aclanthology.org/2022.coling-1.427 | null | null | others | null | Learn2Weight: Parameter Adaptation against Similar-domain Adversarial Attacks | Recent work in black-box adversarial attacks for NLP systems has attracted attention. Prior black-box attacks assume that attackers can observe output labels from target models based on selected inputs. In this work, inspired by adversarial transferability, we propose a new type of black-box NLP adversarial attack that an attacker can choose a similar domain and transfer the adversarial examples to the target domain and cause poor performance in target model. Based on domain adaptation theory, we then propose a defensive strategy, called Learn2Weight, which trains to predict the weight adjustments for target model in order to defense the attack of similar-domain adversarial examples. Using Amazon multi-domain sentiment classification dataset, we empirically show that Learn2Weight model is effective against the attack compared to standard black-box defense methods such as adversarial training and defense distillation. This work contributes to the growing literature on machine learning safety. |
2021.ranlp.xml | Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP 2021) | 2021 | https://aclanthology.org/2021.ranlp-1.113 | null | 1 | hallucination, factuality | Bulgarian | COVID-19 in Bulgarian Social Media: Factuality, Harmfulness, Propaganda, and Framing | With the emergence of the COVID-19 pandemic, the political and the medical aspects of disinformation merged as the problem got elevated to a whole new level to become the first global infodemic. Fighting this infodemic is currently ranked very high on the list of priorities of the World Health Organization, with dangers ranging from promoting fake cures, rumors, and conspiracy theories to spreading xenophobia and panic. With this in mind, we studied how COVID-19 is discussed in Bulgarian social media in terms of factuality, harmfulness, propaganda, and framing. We found that most Bulgarian tweets contain verifiable factual claims, are factually true, are of potential public interest, are not harmful, and are too trivial to fact-check; moreover, zooming into harmful tweets, we found that they spread not only rumors but also panic. We further analyzed articles shared in Bulgarian partisan pro/con-COVID-19 Facebook groups and found that propaganda is more prevalent in skeptical articles, which use doubt, flag waving, and slogans to convey their message; in contrast, concerned ones appeal to emotions, fear, and authority; moreover, skeptical articles frame the issue as one of quality of life, policy, legality, economy, and politics, while concerned articles focus on health & safety. We release our manually and automatically analyzed datasets to enable further research. |
2021.ranlp.xml | Proceedings of the International Conference on Recent Advances in Natural Language Processing (RANLP 2021) | 2021 | https://aclanthology.org/2021.ranlp-1.114 | null | 1 | hallucination, factuality | English, Arabic | A Second Pandemic? Analysis of Fake News about COVID-19 Vaccines in Qatar | While COVID-19 vaccines are finally becoming widely available, a second pandemic that revolves around the circulation of anti-vaxxer “fake news” may hinder efforts to recover from the first one. With this in mind, we performed an extensive analysis of Arabic and English tweets about COVID-19 vaccines, with focus on messages originating from Qatar. We found that Arabic tweets contain a lot of false information and rumors, while English tweets are mostly factual. However, English tweets are much more propagandistic than Arabic ones. In terms of propaganda techniques, about half of the Arabic tweets express doubt, and 1/5 use loaded language, while English tweets are abundant in loaded language, exaggeration, fear, name-calling, doubt, and flag-waving. Finally, in terms of framing, Arabic tweets adopt a health and safety perspective, while in English economic concerns dominate. |
2021.acl.xml | Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) | 2021 | https://aclanthology.org/2021.acl-long.85 | null | null | others | null | Unsupervised Out-of-Domain Detection via Pre-trained Transformers | Deployed real-world machine learning applications are often subject to uncontrolled and even potentially malicious inputs. Those out-of-domain inputs can lead to unpredictable outputs and sometimes catastrophic safety issues. Prior studies on out-of-domain detection require in-domain task labels and are limited to supervised classification scenarios. Our work tackles the problem of detecting out-of-domain samples with only unsupervised in-domain data. We utilize the latent representations of pre-trained transformers and propose a simple yet effective method to transform features across all layers to construct out-of-domain detectors efficiently. Two domain-specific fine-tuning approaches are further proposed to boost detection accuracy. Our empirical evaluations of related methods on two datasets validate that our method greatly improves out-of-domain detection ability in a more general scenario. |
2021.acl.xml | Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) | 2021 | https://aclanthology.org/2021.acl-long.320 | null | null | others | null | Supporting Land Reuse of Former Open Pit Mining Sites using Text Classification and Active Learning | Open pit mines left many regions worldwide inhospitable or uninhabitable. Many sites are left behind in a hazardous or contaminated state, show remnants of waste, or have other restrictions imposed upon them, e.g., for the protection of human or nature. Such information has to be permanently managed in order to reuse those areas in the future. In this work we present and evaluate an automated workflow for supporting the post-mining management of former lignite open pit mines in the eastern part of Germany, where prior to any planned land reuse, aforementioned information has to be acquired to ensure the safety and validity of such an endeavor. Usually, this information is found in expert reports, either in the form of paper documents, or in the best case as digitized unstructured text—all of them in German language. However, due to the size and complexity of these documents, any inquiry is tedious and time-consuming, thereby slowing down or even obstructing the reuse of related areas. Since no training data is available, we employ active learning in order to perform multi-label sentence classification for two categories of restrictions and seven categories of topics. The final system integrates optical character recognition (OCR), active-learning-based text classification, and geographic information system visualization in order to effectively extract, query, and visualize this information for any area of interest. Active learning and text classification results are twofold: Whereas the restriction categories were reasonably accurate (>0.85 F1), the seven topic-oriented categories seemed to be complex even for human annotators and achieved mediocre evaluation scores (<0.70 F1). |
2021.eacl.xml | Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume | 2021 | https://aclanthology.org/2021.eacl-main.232 | null | 0 | general safety, LLM alignment | null | BERTective: Language Models and Contextual Information for Deception Detection | Spotting a lie is challenging but has an enormous potential impact on security as well as private and public safety. Several NLP methods have been proposed to classify texts as truthful or deceptive. In most cases, however, the target texts’ preceding context is not considered. This is a severe limitation, as any communication takes place in context, not in a vacuum, and context can help to detect deception. We study a corpus of Italian dialogues containing deceptive statements and implement deep neural models that incorporate various linguistic contexts. We establish a new state-of-the-art identifying deception and find that not all context is equally useful to the task. Only the texts closest to the target, if from the same speaker (rather than questions by an interlocutor), boost performance. We also find that the semantic information in language models such as BERT contributes to the performance. However, BERT alone does not capture the implicit knowledge of deception cues: its contribution is conditional on the concurrent use of attention to learn cues from BERT’s representations. |
2021.emnlp.xml | Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing | 2021 | https://aclanthology.org/2021.emnlp-main.37 | null | 0 | jailbreaking attacks | null | Automatically Exposing Problems with Neural Dialog Models | Neural dialog models are known to suffer from problems such as generating unsafe and inconsistent responses. Even though these problems are crucial and prevalent, they are mostly manually identified by model designers through interactions. Recently, some research instructs crowdworkers to goad the bots into triggering such problems. However, humans leverage superficial clues such as hate speech, while leaving systematic problems undercover. In this paper, we propose two methods including reinforcement learning to automatically trigger a dialog model into generating problematic responses. We show the effect of our methods in exposing safety and contradiction issues with state-of-the-art dialog models. |
2021.emnlp.xml | Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing | 2021 | https://aclanthology.org/2021.emnlp-main.398 | null | 0 | toxicity, bias | null | Multi-Modal Open-Domain Dialogue | Recent work in open-domain conversational agents has demonstrated that significant improvements in humanness and user preference can be achieved via massive scaling in both pre-training data and model size (Adiwardana et al., 2020; Roller et al., 2020). However, if we want to build agents with human-like abilities, we must expand beyond handling just text. A particularly important topic is the ability to see images and communicate about what is perceived. With the goal of getting humans to engage in multi-modal dialogue, we investigate combining components from state-of-the-art open-domain dialogue agents with those from state-of-the-art vision models. We study incorporating different image fusion schemes and domain-adaptive pre-training and fine-tuning strategies, and show that our best resulting model outperforms strong existing models in multi-modal dialogue while simultaneously performing as well as its predecessor (text-only) BlenderBot (Roller et al., 2020) in text-based conversation. We additionally investigate and incorporate safety components in our final model, and show that such efforts do not diminish model performance with respect to human preference. |
2021.emnlp.xml | Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing | 2021 | https://aclanthology.org/2021.emnlp-main.659 | x | 0 | jailbreaking attacks | null | RAP: Robustness-Aware Perturbations for Defending against Backdoor Attacks on NLP Models | Backdoor attacks, which maliciously control a well-trained model’s outputs of the instances with specific triggers, are recently shown to be serious threats to the safety of reusing deep neural networks (DNNs). In this work, we propose an efficient online defense mechanism based on robustness-aware perturbations. Specifically, by analyzing the backdoor training process, we point out that there exists a big gap of robustness between poisoned and clean samples. Motivated by this observation, we construct a word-based robustness-aware perturbation to distinguish poisoned samples from clean samples to defend against the backdoor attacks on natural language processing (NLP) models. Moreover, we give a theoretical analysis about the feasibility of our robustness-aware perturbation-based defense method. Experimental results on sentiment analysis and toxic detection tasks show that our method achieves better defending performance and much lower computational costs than existing online defense methods. Our code is available at https://github.com/lancopku/RAP. |
2021.woah.xml | Proceedings of the 5th Workshop on Online Abuse and Harms (WOAH 2021) | 2021 | https://aclanthology.org/2021.woah-1.7 | null | 1 | toxicity, bias | nepalese | Offensive Language Detection in Nepali Social Media | Social media texts such as blog posts, comments, and tweets often contain offensive languages including racial hate speech comments, personal attacks, and sexual harassment. Detecting inappropriate use of language is, therefore, of utmost importance for the safety of the users as well as for suppressing hateful conduct and aggression. Existing approaches to this problem are mostly available for resource-rich languages such as English and German. In this paper, we characterize the offensive language in Nepali, a low-resource language, highlighting the challenges that need to be addressed for processing Nepali social media text. We also present experiments for detecting offensive language using supervised machine learning. Besides contributing the first baseline approaches of detecting offensive language in Nepali, we also release human annotated data sets to encourage future research on this crucial topic. |
2021.naacl.xml | Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies | 2021 | https://aclanthology.org/2021.naacl-main.121 | null | 0 | hallucination, factuality | null | How Robust are Fact Checking Systems on Colloquial Claims? | Knowledge is now starting to power neural dialogue agents. At the same time, the risk of misinformation and disinformation from dialogue agents also rises. Verifying the veracity of information from formal sources are widely studied in computational fact checking. In this work, we ask: How robust are fact checking systems on claims in colloquial style? We aim to open up new discussions in the intersection of fact verification and dialogue safety. In order to investigate how fact checking systems behave on colloquial claims, we transfer the styles of claims from FEVER (Thorne et al., 2018) into colloquialism. We find that existing fact checking systems that perform well on claims in formal style significantly degenerate on colloquial claims with the same semantics. Especially, we show that document retrieval is the weakest spot in the system even vulnerable to filler words, such as “yeah” and “you know”. The document recall of WikiAPI retriever (Hanselowski et al., 2018) which is 90.0% on FEVER, drops to 72.2% on the colloquial claims. We compare the characteristics of colloquial claims to those of claims in formal style, and demonstrate the challenging issues in them. |
2021.naacl.xml | Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies | 2021 | https://aclanthology.org/2021.naacl-main.190 | null | 0 | toxicity, bias | null | Detoxifying Language Models Risks Marginalizing Minority Voices | Language models (LMs) must be both safe and equitable to be responsibly deployed in practice. With safety in mind, numerous detoxification techniques (e.g., Dathathri et al. 2020; Krause et al. 2020) have been proposed to mitigate toxic LM generations. In this work, we show that these detoxification techniques hurt equity: they decrease the utility of LMs on language used by marginalized groups (e.g., African-American English and minority identity mentions). In particular, we perform automatic and human evaluations of text generation quality when LMs are conditioned on inputs with different dialects and group identifiers. We find that detoxification makes LMs more brittle to distribution shift, especially on language used by marginalized groups. We identify that these failures stem from detoxification methods exploiting spurious correlations in toxicity datasets. Overall, our results highlight the tension between the controllability and distributional robustness of LMs. |
2021.naacl.xml | Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies | 2021 | https://aclanthology.org/2021.naacl-main.235 | null | 0 | general safety, LLM alignment | null | Bot-Adversarial Dialogue for Safe Conversational Agents | Conversational agents trained on large unlabeled corpora of human interactions will learn patterns and mimic behaviors therein, which include offensive or otherwise toxic behavior. We introduce a new human-and-model-in-the-loop framework for evaluating the toxicity of such models, and compare a variety of existing methods in both the cases of non-adversarial and adversarial users that expose their weaknesses. We then go on to propose two novel methods for safe conversational agents, by either training on data from our new human-and-model-in-the-loop framework in a two-stage system, or ”baking-in” safety to the generative model itself. We find our new techniques are (i) safer than existing models; while (ii) maintaining usability metrics such as engagingness relative to state-of-the-art chatbots. In contrast, we expose serious safety issues in existing standard systems like GPT2, DialoGPT, and BlenderBot. |
2021.findings.xml | Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021 | 2021 | https://aclanthology.org/2021.findings-emnlp.210 | null | 0 | toxicity, bias | null | Challenges in Detoxifying Language Models | Large language models (LM) generate remarkably fluent text and can be efficiently adapted across NLP tasks. Measuring and guaranteeing the quality of generated text in terms of safety is imperative for deploying LMs in the real world; to this end, prior work often relies on automatic evaluation of LM toxicity. We critically discuss this approach, evaluate several toxicity mitigation strategies with respect to both automatic and human evaluation, and analyze consequences of toxicity mitigation in terms of model bias and LM quality. We demonstrate that while basic intervention strategies can effectively optimize previously established automatic metrics on the REALTOXICITYPROMPTS dataset, this comes at the cost of reduced LM coverage for both texts about, and dialects of, marginalized groups. Additionally, we find that human raters often disagree with high automatic toxicity scores after strong toxicity reduction interventions—highlighting further the nuances involved in careful evaluation of LM toxicity. |
2020.starsem.xml | Proceedings of the Ninth Joint Conference on Lexical and Computational Semantics | 2020 | https://aclanthology.org/2020.starsem-1.19 | null | null | others | null | Large Scale Author Obfuscation Using Siamese Variational Auto-Encoder: The SiamAO System | Author obfuscation is the task of masking the author of a piece of text, with applications in privacy. Recent advances in deep neural networks have boosted author identification performance making author obfuscation more challenging. Existing approaches to author obfuscation are largely heuristic. Obfuscation can, however, be thought of as the construction of adversarial examples to attack author identification, suggesting that the deep learning architectures used for adversarial attacks could have application here. Current architectures are proposed to construct adversarial examples against classification-based models, which in author identification would exclude the high-performing similarity-based models employed when facing large number of authorial classes. In this paper, we propose the first deep learning architecture for constructing adversarial examples against similarity-based learners, and explore its application to author obfuscation. We analyse the output from both success in obfuscation and language acceptability, as well as comparing the performance with some common baselines, and showing promising results in finding a balance between safety and soundness of the perturbed texts. |
2020.emnlp.xml | Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) | 2020 | https://aclanthology.org/2020.emnlp-main.656 | x | 0 | toxicity, bias | null | Queens are Powerful too: Mitigating Gender Bias in Dialogue Generation | Social biases present in data are often directly reflected in the predictions of models trained on that data. We analyze gender bias in dialogue data, and examine how this bias is not only replicated, but is also amplified in subsequent generative chit-chat dialogue models. We measure gender bias in six existing dialogue datasets before selecting the most biased one, the multi-player text-based fantasy adventure dataset LIGHT, as a testbed for bias mitigation techniques. We consider three techniques to mitigate gender bias: counterfactual data augmentation, targeted data collection, and bias controlled training. We show that our proposed techniques mitigate gender bias by balancing the genderedness of generated dialogue utterances, and find that they are particularly effective in combination. We evaluate model performance with a variety of quantitative methods—including the quantity of gendered words, a dialogue safety classifier, and human assessments—all of which show that our models generate less gendered, but equally engaging chit-chat responses. |
2020.acl.xml | Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics | 2020 | https://aclanthology.org/2020.acl-main.188 | null | null | others | null | Calibrating Structured Output Predictors for Natural Language Processing | We address the problem of calibrating prediction confidence for output entities of interest in natural language processing (NLP) applications. It is important that NLP applications such as named entity recognition and question answering produce calibrated confidence scores for their predictions, especially if the applications are to be deployed in a safety-critical domain such as healthcare. However the output space of such structured prediction models are often too large to directly adapt binary or multi-class calibration methods. In this study, we propose a general calibration scheme for output entities of interest in neural network based structured prediction models. Our proposed method can be used with any binary class calibration scheme and a neural network model. Additionally, we show that our calibration method can also be used as an uncertainty-aware, entity-specific decoding step to improve the performance of the underlying model at no additional training cost or data requirements. We show that our method outperforms current calibration techniques for Named Entity Recognition, Part-of-speech tagging and Question Answering systems. We also observe an improvement in model performance from our decoding step across several tasks and benchmark datasets. Our method improves the calibration and model performance on out-of-domain test scenarios as well. |
2020.alw.xml | Proceedings of the Fourth Workshop on Online Abuse and Harms | 2020 | https://aclanthology.org/2020.alw-1.11 | null | 0 | general safety, LLM alignment | null | Enhancing the Identification of Cyberbullying through Participant Roles | Cyberbullying is a prevalent social problem that inflicts detrimental consequences to the health and safety of victims such as psychological distress, anti-social behaviour, and suicide. The automation of cyberbullying detection is a recent but widely researched problem, with current research having a strong focus on a binary classification of bullying versus non-bullying. This paper proposes a novel approach to enhancing cyberbullying detection through role modeling. We utilise a dataset from ASKfm to perform multi-class classification to detect participant roles (e.g. victim, harasser). Our preliminary results demonstrate promising performance including 0.83 and 0.76 of F1-score for cyberbullying and role classification respectively, outperforming baselines. |
2020.osact.xml | Proceedings of the 4th Workshop on Open-Source Arabic Corpora and Processing Tools, with a Shared Task on Offensive Language Detection | 2020 | https://aclanthology.org/2020.osact-1.6 | null | 1 | toxicity, bias | Modern Standard Arabic, Gulf, Egyptian, Maghrebi, Levantine | Understanding and Detecting Dangerous Speech in Social Media | Social media communication has become a significant part of daily activity in modern societies. For this reason, ensuring safety in social media platforms is a necessity. Use of dangerous language such as physical threats in online environments is a somewhat rare, yet remains highly important. Although several works have been performed on the related issue of detecting offensive and hateful language, dangerous speech has not previously been treated in any significant way. Motivated by these observations, we report our efforts to build a labeled dataset for dangerous speech. We also exploit our dataset to develop highly effective models to detect dangerous content. Our best model performs at 59.60% macro F1, significantly outperforming a competitive baseline. |
2020.clinicalnlp.xml | Proceedings of the 3rd Clinical Natural Language Processing Workshop | 2020 | https://aclanthology.org/2020.clinicalnlp-1.29 | x | 0 | hallucination, factuality | null | Utilizing Multimodal Feature Consistency to Detect Adversarial Examples on Clinical Summaries | Recent studies have shown that adversarial examples can be generated by applying small perturbations to the inputs such that the well- trained deep learning models will misclassify. With the increasing number of safety and security-sensitive applications of deep learn- ing models, the robustness of deep learning models has become a crucial topic. The robustness of deep learning models for health- care applications is especially critical because the unique characteristics and the high financial interests of the medical domain make it more sensitive to adversarial attacks. Among the modalities of medical data, the clinical summaries have higher risks to be attacked because they are generated by third-party companies. As few works studied adversarial threats on clinical summaries, in this work we first apply adversarial attack to clinical summaries of electronic health records (EHR) to show the text-based deep learning systems are vulnerable to adversarial examples. Secondly, benefiting from the multi-modality of the EHR dataset, we propose a novel defense method, MATCH (Multimodal feATure Consistency cHeck), which leverages the consistency between multiple modalities in the data to defend against adversarial examples on a single modality. Our experiments demonstrate the effectiveness of MATCH on a hospital readmission prediction task comparing with baseline methods. |
2020.nlpmc.xml | Proceedings of the First Workshop on Natural Language Processing for Medical Conversations | 2020 | https://aclanthology.org/2020.nlpmc-1.1 | null | null | others | null | Methods for Extracting Information from Messages from Primary Care Providers to Specialists | Electronic consult (eConsult) systems allow specialists more flexibility to respond to referrals more efficiently, thereby increasing access in under-resourced healthcare settings like safety net systems. Understanding the usage patterns of eConsult system is an important part of improving specialist efficiency. In this work, we develop and apply classifiers to a dataset of eConsult questions from primary care providers to specialists, classifying the messages for how they were triaged by the specialist office, and the underlying type of clinical question posed by the primary care provider. We show that pre-trained transformer models are strong baselines, with improving performance from domain-specific training and shared representations. |
2020.lrec.xml | Proceedings of the Twelfth Language Resources and Evaluation Conference | 2020 | https://aclanthology.org/2020.lrec-1.245 | null | null | others | null | Semantic Annotation for Improved Safety in Construction Work | Risk management is a vital activity to ensure employee safety in construction projects. Various documents provide important supporting evidence, including details of previous incidents, consequences and mitigation strategies. Potential hazards may depend on a complex set of project-specific attributes, including activities undertaken, location, equipment used, etc. However, finding evidence about previous projects with similar attributes can be problematic, since information about risks and mitigations is usually hidden within and may be dispersed across a range of different free text documents. Automatic named entity recognition (NER), which identifies mentions of concepts in free text documents, is the first stage in structuring knowledge contained within them. While developing NER methods generally relies on annotated corpora, we are not aware of any such corpus targeted at concepts relevant to construction safety. In response, we have designed a novel named entity annotation scheme and associated guidelines for this domain, which covers hazards, consequences, mitigation strategies and project attributes. Four health and safety experts used the guidelines to annotate a total of 600 sentences from accident reports; an average inter-annotator agreement rate of 0.79 F-Score shows that our work constitutes an important first step towards developing tools for detailed semantic analysis of construction safety documents. |
2020.lrec.xml | Proceedings of the Twelfth Language Resources and Evaluation Conference | 2020 | https://aclanthology.org/2020.lrec-1.703 | null | null | others | null | Must Children be Vaccinated or not? Annotating Modal Verbs in the Vaccination Debate | In this paper we analyze the use of modal verbs in a corpus of texts related to the vaccination debate. Broadly speaking, the vaccination debate centers around whether vaccination is safe, and whether it is morally acceptable to enforce mandatory vaccination. In order to successfully intervene and curb the spread of preventable diseases due to low vaccination rates, health practitioners need to be adequately informed on public perception of the safety and necessity of vaccines. Public perception can relate to the strength of conviction that an individual may have towards a proposition (e.g. ‘one must vaccinate’ versus ‘one should vaccinate’), as well as qualify the type of proposition, be it related to morality (‘government should not interfere in my personal choice’) or related to possibility (‘too many vaccines at once could hurt my child’). Text mining and analysis of modal auxiliaries are economically viable means of gaining insights into these perspectives, particularly on a large scale due to the widespread use of social media and blogs as vehicles of communication. |
2020.lrec.xml | Proceedings of the Twelfth Language Resources and Evaluation Conference | 2020 | https://aclanthology.org/2020.lrec-1.714 | null | null | others | null | Dataset and Enhanced Model for Eligibility Criteria-to-SQL Semantic Parsing | Clinical trials often require that patients meet eligibility criteria (e.g., have specific conditions) to ensure the safety and the effectiveness of studies. However, retrieving eligible patients for a trial from the electronic health record (EHR) database remains a challenging task for clinicians since it requires not only medical knowledge about eligibility criteria, but also an adequate understanding of structured query language (SQL). In this paper, we introduce a new dataset that includes the first-of-its-kind eligibility-criteria corpus and the corresponding queries for criteria-to-sql (Criteria2SQL), a task translating the eligibility criteria to executable SQL queries. Compared to existing datasets, the queries in the dataset here are derived from the eligibility criteria of clinical trials and include Order-sensitive, Counting-based, and Boolean-type cases which are not seen before. In addition to the dataset, we propose a novel neural semantic parser as a strong baseline model. Extensive experiments show that the proposed parser outperforms existing state-of-the-art general-purpose text-to-sql models while highlighting the challenges presented by the new dataset. The uniqueness and the diversity of the dataset leave a lot of research opportunities for future improvement. |
2020.lrec.xml | Proceedings of the Twelfth Language Resources and Evaluation Conference | 2020 | https://aclanthology.org/2020.lrec-1.747 | x | 0 | hallucination, factuality | null | A Survey on Natural Language Processing for Fake News Detection | Fake news detection is a critical yet challenging problem in Natural Language Processing (NLP). The rapid rise of social networking platforms has not only yielded a vast increase in information accessibility but has also accelerated the spread of fake news. Thus, the effect of fake news has been growing, sometimes extending to the offline world and threatening public safety. Given the massive amount of Web content, automatic fake news detection is a practical NLP problem useful to all online content providers, in order to reduce the human time and effort to detect and prevent the spread of fake news. In this paper, we describe the challenges involved in fake news detection and also describe related tasks. We systematically review and compare the task formulations, datasets and NLP solutions that have been developed for this task, and also discuss the potentials and limitations of them. Based on our insights, we outline promising research directions, including more fine-grained, detailed, fair, and practical detection models. We also highlight the difference between fake news detection and other related tasks, and the importance of NLP solutions for fake news detection. |
2020.lrec.xml | Proceedings of the Twelfth Language Resources and Evaluation Conference | 2020 | https://aclanthology.org/2020.lrec-1.794 | null | 0 | general safety, LLM alignment | null | The SAFE-T Corpus: A New Resource for Simulated Public Safety Communications | We introduce a new resource, the SAFE-T (Speech Analysis for Emergency Response Technology) Corpus, designed to simulate first-responder communications by inducing high vocal effort and urgent speech with situational background noise in a game-based collection protocol. Linguistic Data Consortium developed the SAFE-T Corpus to support the NIST (National Institute of Standards and Technology) OpenSAT (Speech Analytic Technologies) evaluation series, whose goal is to advance speech analytic technologies including automatic speech recognition, speech activity detection and keyword search in multiple domains including simulated public safety communications data. The corpus comprises over 300 hours of audio from 115 unique speakers engaged in a collaborative problem-solving activity representative of public safety communications in terms of speech content, noise types and noise levels. Portions of the corpus have been used in the OpenSAT 2019 evaluation and the full corpus will be published in the LDC catalog. We describe the design and implementation of the SAFE-T Corpus collection, discuss the approach of capturing spontaneous speech from study participants through game-based speech collection, and report on the collection results including several challenges associated with the collection. |
2020.wnut.xml | Proceedings of the Sixth Workshop on Noisy User-generated Text (W-NUT 2020) | 2020 | https://aclanthology.org/2020.wnut-1.55 | null | 0 | hallucination, factuality | null | CSECU-DSG at WNUT-2020 Task 2: Exploiting Ensemble of Transfer Learning and Hand-crafted Features for Identification of Informative COVID-19 English Tweets | COVID-19 pandemic has become the trending topic on twitter and people are interested in sharing diverse information ranging from new cases, healthcare guidelines, medicine, and vaccine news. Such information assists the people to be updated about the situation as well as beneficial for public safety personnel for decision making. However, the informal nature of twitter makes it challenging to refine the informative tweets from the huge tweet streams. To address these challenges WNUT-2020 introduced a shared task focusing on COVID-19 related informative tweet identification. In this paper, we describe our participation in this task. We propose a neural model that adopts the strength of transfer learning and hand-crafted features in a unified architecture. To extract the transfer learning features, we utilize the state-of-the-art pre-trained sentence embedding model BERT, RoBERTa, and InferSent, whereas various twitter characteristics are exploited to extract the hand-crafted features. Next, various feature combinations are utilized to train a set of multilayer perceptron (MLP) as the base-classifier. Finally, a majority voting based fusion approach is employed to determine the informative tweets. Our approach achieved competitive performance and outperformed the baseline by 7% (approx.). |
D19.xml | Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) | 2019 | https://aclanthology.org/D19-5107 | null | null | others | null | Complaint Analysis and Classification for Economic and Food Safety | Governmental institutions are employing artificial intelligence techniques to deal with their specific problems and exploit their huge amounts of both structured and unstructured information. In particular, natural language processing and machine learning techniques are being used to process citizen feedback. In this paper, we report on the use of such techniques for analyzing and classifying complaints, in the context of the Portuguese Economic and Food Safety Authority. Grounded in its operational process, we address three different classification problems: target economic activity, implied infraction severity level, and institutional competence. We show promising results obtained using feature-based approaches and traditional classifiers, with accuracy scores above 70%, and analyze the shortcomings of our current results and avenues for further improvement, taking into account the intended use of our classifiers in helping human officers to cope with thousands of yearly complaints. |
S19.xml | Proceedings of the Eighth Joint Conference on Lexical and Computational Semantics (* | 2019 | https://aclanthology.org/S19-1032 | null | null | others | null | Generating Animations from Screenplays | Automatically generating animation from natural language text finds application in a number of areas e.g. movie script writing, instructional videos, and public safety. However, translating natural language text into animation is a challenging task. Existing text-to-animation systems can handle only very simple sentences, which limits their applications. In this paper, we develop a text-to-animation system which is capable of handling complex sentences. We achieve this by introducing a text simplification step into the process. Building on an existing animation generation system for screenwriting, we create a robust NLP pipeline to extract information from screenplays and map them to the system’s knowledge base. We develop a set of linguistic transformation rules that simplify complex sentences. Information extracted from the simplified sentences is used to generate a rough storyboard and video depicting the text. Our sentence simplification module outperforms existing systems in terms of BLEU and SARI metrics. We further evaluated our system via a user study: 68% participants believe that our system generates reasonable animation from input screenplays. |
W19.xml | Proceedings of the Society for Computation in Linguistics ( | 2019 | https://aclanthology.org/W19-5036 | null | null | others | null | Improving classification of Adverse Drug Reactions through Using Sentiment Analysis and Transfer Learning | The availability of large-scale and real-time data on social media has motivated research into adverse drug reactions (ADRs). ADR classification helps to identify negative effects of drugs, which can guide health professionals and pharmaceutical companies in making medications safer and advocating patients’ safety. Based on the observation that in social media, negative sentiment is frequently expressed towards ADRs, this study presents a neural model that combines sentiment analysis with transfer learning techniques to improve ADR detection in social media postings. Our system is firstly trained to classify sentiment in tweets concerning current affairs, using the SemEval17-task4A corpus. We then apply transfer learning to adapt the model to the task of detecting ADRs in social media postings. We show that, in combination with rich representations of words and their contexts, transfer learning is beneficial, especially given the large degree of vocabulary overlap between the current affairs posts in the SemEval17-task4A corpus and posts about ADRs. We compare our results with previous approaches, and show that our model can outperform them by up to 3% F-score. |
C18.xml | Proceedings of the 27th International Conference on Computational Linguistics | 2018 | https://aclanthology.org/C18-1212 | null | null | others | null | Real-time Change Point Detection using On-line Topic Models | Detecting changes within an unfolding event in real time from news articles or social media enables to react promptly to serious issues in public safety, public health or natural disasters. In this study, we use on-line Latent Dirichlet Allocation (LDA) to model shifts in topics, and apply on-line change point detection (CPD) algorithms to detect when significant changes happen. We describe an on-line Bayesian change point detection algorithm that we use to detect topic changes from on-line LDA output. Extensive experiments on social media data and news articles show the benefits of on-line LDA versus standard LDA, and of on-line change point detection compared to off-line algorithms. This yields F-scores up to 52% on the detection of significant real-life changes from these document streams. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.