{ "paper_id": "P18-1024", "header": { "generated_with": "S2ORC 1.0.0", "date_generated": "2023-01-19T08:41:13.219462Z" }, "title": "LinkNBed: Multi-Graph Representation Learning with Entity Linkage", "authors": [ { "first": "Rakshit", "middle": [], "last": "Trivedi", "suffix": "", "affiliation": {}, "email": "rstrivedi@gatech.edu" }, { "first": "Christos", "middle": [], "last": "Faloutsos", "suffix": "", "affiliation": {}, "email": "" }, { "first": "Bunyamin", "middle": [], "last": "Sisman", "suffix": "", "affiliation": {}, "email": "bunyamis@amazon.com.work" }, { "first": "Hongyuan", "middle": [], "last": "Zha", "suffix": "", "affiliation": {}, "email": "" }, { "first": "Jun", "middle": [], "last": "Ma", "suffix": "", "affiliation": {}, "email": "" }, { "first": "Xin", "middle": [ "Luna" ], "last": "Dong", "suffix": "", "affiliation": {}, "email": "" } ], "year": "", "venue": null, "identifiers": {}, "abstract": "Knowledge graphs have emerged as an important model for studying complex multirelational data. This has given rise to the construction of numerous large scale but incomplete knowledge graphs encoding information extracted from various resources. An effective and scalable approach to jointly learn over multiple graphs and eventually construct a unified graph is a crucial next step for the success of knowledge-based inference for many downstream applications. To this end, we propose LinkNBed, a deep relational learning framework that learns entity and relationship representations across multiple graphs. We identify entity linkage across graphs as a vital component to achieve our goal. We design a novel objective that leverage entity linkage and build an efficient multi-task training procedure. Experiments on link prediction and entity linkage demonstrate substantial improvements over the state-ofthe-art relational learning approaches.", "pdf_parse": { "paper_id": "P18-1024", "_pdf_hash": "", "abstract": [ { "text": "Knowledge graphs have emerged as an important model for studying complex multirelational data. This has given rise to the construction of numerous large scale but incomplete knowledge graphs encoding information extracted from various resources. An effective and scalable approach to jointly learn over multiple graphs and eventually construct a unified graph is a crucial next step for the success of knowledge-based inference for many downstream applications. To this end, we propose LinkNBed, a deep relational learning framework that learns entity and relationship representations across multiple graphs. We identify entity linkage across graphs as a vital component to achieve our goal. We design a novel objective that leverage entity linkage and build an efficient multi-task training procedure. Experiments on link prediction and entity linkage demonstrate substantial improvements over the state-ofthe-art relational learning approaches.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Abstract", "sec_num": null } ], "body_text": [ { "text": "Reasoning over multi-relational data is a key concept in Artificial Intelligence and knowledge graphs have appeared at the forefront as an effective tool to model such multi-relational data. Knowledge graphs have found increasing importance due to its wider range of important applications such as information retrieval (Dalton et al., 2014) , natural language processing (Gabrilovich and Markovitch, 2009) , recommender systems (Catherine and Cohen, 2016) , question-answering (Cui et al., 2017) and many more. This has led to the increased efforts in constructing numerous large-scale Knowledge Bases (e.g. Freebase (Bollacker et al., 2008) , DBpedia (Auer et al., 2007) , Google's Knowledge graph (Dong et al., 2014) , Yago (Suchanek et al., 2007) and NELL (Carlson et al., 2010) ), that can cater to these applications, by representing information available on the web in relational format.", "cite_spans": [ { "start": 320, "end": 341, "text": "(Dalton et al., 2014)", "ref_id": "BIBREF12" }, { "start": 372, "end": 406, "text": "(Gabrilovich and Markovitch, 2009)", "ref_id": "BIBREF17" }, { "start": 429, "end": 456, "text": "(Catherine and Cohen, 2016)", "ref_id": "BIBREF9" }, { "start": 478, "end": 496, "text": "(Cui et al., 2017)", "ref_id": "BIBREF11" }, { "start": 618, "end": 642, "text": "(Bollacker et al., 2008)", "ref_id": "BIBREF1" }, { "start": 653, "end": 672, "text": "(Auer et al., 2007)", "ref_id": "BIBREF0" }, { "start": 700, "end": 719, "text": "(Dong et al., 2014)", "ref_id": "BIBREF13" }, { "start": 727, "end": 750, "text": "(Suchanek et al., 2007)", "ref_id": "BIBREF37" }, { "start": 760, "end": 782, "text": "(Carlson et al., 2010)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "All knowledge graphs share common drawback of incompleteness and sparsity and hence most existing relational learning techniques focus on using observed triplets in an incomplete graph to infer unobserved triplets for that graph (Nickel et al., 2016a) . Neural embedding techniques that learn vector space representations of entities and relationships have achieved remarkable success in this task. However, these techniques only focus on learning from a single graph. In addition to incompleteness property, these knowledge graphs also share a set of overlapping entities and relationships with varying information about them. This makes a compelling case to design a technique that can learn over multiple graphs and eventually aid in constructing a unified giant graph out of them. While research on learning representations over single graph has progressed rapidly in recent years (Nickel et al., 2011; Dong et al., 2014; Trouillon et al., 2016; Bordes et al., 2013; Xiao et al., 2016; Yang et al., 2015) , there is a conspicuous lack of principled approach to tackle the unique challenges involved in learning across multiple graphs.", "cite_spans": [ { "start": 229, "end": 251, "text": "(Nickel et al., 2016a)", "ref_id": "BIBREF29" }, { "start": 885, "end": 906, "text": "(Nickel et al., 2011;", "ref_id": "BIBREF31" }, { "start": 907, "end": 925, "text": "Dong et al., 2014;", "ref_id": "BIBREF13" }, { "start": 926, "end": 949, "text": "Trouillon et al., 2016;", "ref_id": "BIBREF40" }, { "start": 950, "end": 970, "text": "Bordes et al., 2013;", "ref_id": "BIBREF3" }, { "start": 971, "end": 989, "text": "Xiao et al., 2016;", "ref_id": "BIBREF42" }, { "start": 990, "end": 1008, "text": "Yang et al., 2015)", "ref_id": "BIBREF43" } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "One approach to multi-graph representation learning could be to first solve graph alignment problem to merge the graphs and then use existing relational learning methods on merged graph. Unfortunately, graph alignment is an important but still unsolved problem and there exist several techniques addressing its challenges (Liu and Yang, 2016; Pershina et al., 2015; Koutra et al., 2013; Buneman and Staworko, 2016) in limited settings.", "cite_spans": [ { "start": 322, "end": 342, "text": "(Liu and Yang, 2016;", "ref_id": "BIBREF27" }, { "start": 343, "end": 365, "text": "Pershina et al., 2015;", "ref_id": "BIBREF32" }, { "start": 366, "end": 386, "text": "Koutra et al., 2013;", "ref_id": "BIBREF23" }, { "start": 387, "end": 414, "text": "Buneman and Staworko, 2016)", "ref_id": "BIBREF5" } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "The key challenges for the graph alignment problem emanate from the fact that the real world data are noisy and intricate in nature. The noisy or sparse data make it difficult to learn robust alignment features, and data abundance leads to computational challenges due to the combinatorial permutations needed for alignment. These challenges are compounded in multi-relational settings due to heterogeneous nodes and edges in such graphs.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "Recently, deep learning has shown significant impact in learning useful information over noisy, large-scale and heterogeneous graph data (Rossi et al., 2017) . We, therefore, posit that combining graph alignment task with deep representation learning across multi-relational graphs has potential to induce a synergistic effect on both tasks. Specifically, we identify that a key component of graph alignment process-entity linkage-also plays a vital role in learning across graphs. For instance, the embeddings learned over two knowledge graphs for an actor should be closer to one another compared to the embeddings of all the other entities. Similarly, the entities that are already aligned together across the two graphs should produce better embeddings due to the shared context and data. To model this phenomenon, we propose LinkNBed, a novel deep learning framework that jointly performs representation learning and graph linkage task. To achieve this, we identify key challenges involved in the learning process and make the following contributions to address them:", "cite_spans": [ { "start": 137, "end": 157, "text": "(Rossi et al., 2017)", "ref_id": null } ], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "\u2022 We propose novel and principled approach towards jointly learning entity representations and entity linkage. The novelty of our framework stems from its ability to support linkage task across heterogeneous types of entities.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "\u2022 We devise a graph-independent inductive framework that learns functions to capture contextual information for entities and relations. It combines the structural and semantic information in individual graphs for joint inference in a principled manner.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "\u2022 Labeled instances (specifically positive instances for linkage task) are typically very sparse and hence we design a novel multi-task loss function where entity linkage task is tackled in robust manner across various learning scenarios such as learning only with unlabeled instances or only with negative instances.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "\u2022 We design an efficient training procedure to perform joint training in linear time in the number of triples. We demonstrate superior performance of our method on two datasets curated from Freebase and IMDB against stateof-the-art neural embedding methods.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Introduction", "sec_num": "1" }, { "text": "A knowledge graph G comprises of set of facts represented as triplets (e s , r, e o ) denoting the relationship r between subject entity e s and object entity e o . Associated to this knowledge graph, we have a set of attributes that describe observed characteristics of an entity. Attributes are represented as set of key-value pairs for each entity and an attribute can have null (missing) value for an entity. We follow Open World Assumption -triplets not observed in knowledge graph are considered to be missing but not false. We assume that there are no duplicate triplets or self-loops.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Knowledge Graph Representation", "sec_num": "2.1" }, { "text": "Definition. Given a collection of knowledge graphs G, Multi-Graph Relational Learning refers to the the task of learning information rich representations of entities and relationships across graphs.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Multi-Graph Relational Learning", "sec_num": "2.2" }, { "text": "The learned embeddings can further be used to infer new knowledge in the form of link prediction or learn new labels in the form of entity linkage. We motivate our work with the setting of two knowledge graphs where given two graphs G 1 , G 2 \u2208 G, the task is to match an entity e G 1 \u2208 G 1 to an entity e G 2 \u2208 G 2 if they represent the same real-world entity. We discuss a straightforward extension of this setting to more than two graphs in Section 7.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Multi-Graph Relational Learning", "sec_num": "2.2" }, { "text": "Notations. Let X and Y represent realization of two such knowledge graphs extracted from two different sources. Let n X e and n Y e represent number of entities in X and Y respectively. Similarly, n X r and n Y r represent number of relations in X and Y . We combine triplets from both X and Y to obtain set of all observed triplets D = {(e s , r, e o ) p } P p=1 where P is total number of available records across from both graphs. Let E and R be the set of all entities and all relations in D respectively. Let |E| = n and |R| = m. In addition to D, we also have set of linkage labels L for entities between X and Y . Each record in L is represented as triplet (e X \u2208 X, e Y \u2208 Y , l \u2208 {0, 1}) where l = 1 when the entities are matched and l = 0 otherwise.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Multi-Graph Relational Learning", "sec_num": "2.2" }, { "text": "We present a novel inductive multi-graph relational learning framework that learns a set of aggregator functions capable of ingesting various contextual information for both entities and relationships in multi-relational graph. These functions encode the ingested structural and semantic information into low-dimensional entity and relation embeddings. Further, we use these representations to learn a relational score function that computes how two entities are likely to be connected in a particular relationship. The key idea behind this formulation is that when a triplet is observed, the relationship between the two entities can be explained using various contextual information such as local neighborhood features of both entities, attribute features of both entities and type information of the entities which participate in that relationship.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Proposed Method: LinkNBed", "sec_num": "3" }, { "text": "We outline two key insights for establishing the relationships between embeddings of the entities over multiple graphs in our framework: Insight 1 (Embedding Similarity): If the two entities e X \u2208 X and e Y \u2208 Y represent the same real-world entity then their embeddings e X and e Y will be close to each other. Insight 2 (Semantic Replacement): For a given triplet t = (e s , r, e o ) \u2208 X, denote g(t) as the function that computes a relational score for t using entity and relation embeddings. If there exists a matching entity e s \u2208 Y for e s \u2208 X, denote t = (e s , r, e o ) obtained after replacing e s with e s . In this case, g(t) \u223c g(t ) i.e. score of triplets t and t will be similar.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Proposed Method: LinkNBed", "sec_num": "3" }, { "text": "For a triplet (e s , r, e o ) \u2208 D, we describe encoding mechanism of LinkNBed as three-layered architecture that computes the final output representations of z r , z e s , z e o for the given triplet. Figure 1 provides an overview of LinkNBed architecture and we describe the three steps below:", "cite_spans": [], "ref_spans": [ { "start": 201, "end": 209, "text": "Figure 1", "ref_id": "FIGREF0" } ], "eq_spans": [], "section": "Proposed Method: LinkNBed", "sec_num": "3" }, { "text": "Entities, Relations, Types and Attributes are first encoded in its basic vector representations. We use these basic representations to derive more complex contextual embeddings further. Entities, Relations and Types. The embedding vectors corresponding to these three components are learned as follows:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Atomic Layer", "sec_num": "3.1" }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "v e s = f (W E e s ) v e o = f (W E e o )", "eq_num": "(1)" } ], "section": "Atomic Layer", "sec_num": "3.1" }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "v r = f (W R r) v t = f (W T t)", "eq_num": "(2)" } ], "section": "Atomic Layer", "sec_num": "3.1" }, { "text": "where v e s ,v e o \u2208 R d . e s , e o \u2208 R n are \"one-hot\" representations of e s and e o respectively. v r \u2208 R k and r \u2208 R m is \"one-hot\" representation of r. v t \u2208 R q and t \u2208 R z is \"one-hot\" representation of t . W E \u2208 R d\u00d7n , W R \u2208 R k\u00d7m and W T \u2208 R q\u00d7z are the entity, relation and type embedding matrices respectively. f is a nonlinear activation function (Relu in our case). W E , W R and W T can be initialized randomly or using pre-trained word embeddings or vector compositions based on name phrases of components (Socher et al., 2013) .", "cite_spans": [ { "start": 523, "end": 544, "text": "(Socher et al., 2013)", "ref_id": "BIBREF36" } ], "ref_spans": [], "eq_spans": [], "section": "Atomic Layer", "sec_num": "3.1" }, { "text": "Attributes. For a given attribute a represented as key-value pair, we use paragraph2vec (Le and Mikolov, 2014) type of embedding network to learn attribute embedding. Specifically, we represent attribute embedding vector as:", "cite_spans": [ { "start": 88, "end": 110, "text": "(Le and Mikolov, 2014)", "ref_id": "BIBREF24" } ], "ref_spans": [], "eq_spans": [], "section": "Atomic Layer", "sec_num": "3.1" }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "a = f (W key a key + W val a val )", "eq_num": "(3)" } ], "section": "Atomic Layer", "sec_num": "3.1" }, { "text": "where a \u2208 R y , a key \u2208 R u and a val \u2208 R v . W key \u2208 R y\u00d7u and W val \u2208 R y\u00d7v . a key will be \"one-hot\" vector and a val will be feature vector.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Atomic Layer", "sec_num": "3.1" }, { "text": "Note that the dimensions of the embedding vectors do not necessarily need to be the same.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Atomic Layer", "sec_num": "3.1" }, { "text": "While the entity and relationship embeddings described above help to capture very generic latent features, embeddings can be further enriched to capture structural information, attribute information and type information to better explain the existence of a fact. Such information can be modeled as context of nodes and edges in the graph. To this end, we design the following canonical aggregator function that learns various contextual information by aggregating over relevant embedding vectors:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "c(z) = AGG({z , \u2200z \u2208 C(z)})", "eq_num": "(4)" } ], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "where c(z) is the vector representation of the aggregated contextual information for component z.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "Here, component z can be either an entity or a relation. C(z) is the set of components in the context of z and z correspond to the vector embeddings of those components. AGG is the aggregator function which can take many forms such Mean, Max, Pooling or more complex LSTM based aggregators. It is plausible that different components in a context may have varied impact on the component for which the embedding is being learned. To account for this, we employ a soft attention mechanism where we learn attention coefficients to weight components based on their impact before aggregating them. We modify Eq. 4 as:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "c(z) = AGG(q(z) * {z , \u2200z \u2208 C(z)})", "eq_num": "(5)" } ], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "where", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "q(z) = exp(\u03b8 z ) z \u2208C(z) exp(\u03b8 z )", "eq_num": "(6)" } ], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "and \u03b8 z 's are the parameters of attention model. Following contextual information is modeled in our framework:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "Entity Neighborhood Context N c (e) \u2208 R d .", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "Given a triplet (e s , r, e o ), the neighborhood context for an entity e s will be the nodes located near e s other than the node e o . This will capture the effect of local neighborhood in the graph surrounding e s that drives e s to participate in fact (e s , r, e o ). We use Mean as aggregator function. As there can be large number of neighbors, we collect the neighborhood set for each entity as a pre-processing step using a random walk method. Specifically, given a node e, we run k rounds of random-walks of length l following (Hamilton et al., 2017) and create set N (e) by adding all unique nodes visited across these walks. This context can be similarly computed for object entity.", "cite_spans": [ { "start": 537, "end": 560, "text": "(Hamilton et al., 2017)", "ref_id": "BIBREF19" } ], "ref_spans": [], "eq_spans": [], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "Entity Attribute Context A c (e) \u2208 R y . For an entity e, we collect all attribute embeddings for e obtained from Atomic Layer and learn aggregated information over them using Max operator given in Eq. 4.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "Relation Type Context T c (r) \u2208 R q . We use type context for relation embedding i.e. for a given relationship r, this context aims at capturing the effect of type of entities that have participated in this relationship. For a given triplet (e s , r, e o ), type context for relationship r is computed by aggregation with mean over type embeddings corresponding to the context of r. Appendix C provides specific forms of contextual information.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Contextual Layer", "sec_num": "3.2" }, { "text": "Having computed the atomic and contextual embeddings for a triplet (e s , r, e o ), we obtain the final embedded representations of entities and relation in the triplet using the following formulation:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Representation Layer", "sec_num": "3.3" }, { "text": "z e s = \u03c3( W 1 v e s Subject Entity Embedding + W 2 N c (e s ) Neighborhood Context + W 3 A c (e s )) Subject Entity Attributes (7) z e o = \u03c3( W 1 v e o Object Entity Embedding + W 2 N c (e o ) Neighborhood Context + W 3 A c (e o ))", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Representation Layer", "sec_num": "3.3" }, { "text": "Object Entity Attributes (8)", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Representation Layer", "sec_num": "3.3" }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "z r = \u03c3( W 4 v r Relation Embedding + W 5 T c (r)) Entity Type Context", "eq_num": "(9)" } ], "section": "Representation Layer", "sec_num": "3.3" }, { "text": "where", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Representation Layer", "sec_num": "3.3" }, { "text": "W 1 , W 2 \u2208 R d\u00d7d , W 3 \u2208 R d\u00d7y , W 4 \u2208 R d\u00d7k and W 5 \u2208 R d\u00d7q . \u03c3 is nonlinear activation function -generally Tanh or Relu.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Representation Layer", "sec_num": "3.3" }, { "text": "Following is the rationale for our formulation: An entity's representation can be enriched by encoding information about the local neighborhood features and attribute information associated with the entity in addition to its own latent features. Parameters W 1 , W 2 , W 3 learn to capture these different aspects and map them into the entity embedding space. Similarly, a relation's representation can be enriched by encoding information about entity types that participate in that relationship in addition to its own latent features. Parameters W 4 , W 5 learn to capture these aspects and map them into the relation embedding space. Further, as the ultimate goal is to jointly learn over multiple graphs, shared parameterization in our model facilitate the propagation of information across graphs thereby making it a graph-independent inductive model. The flexibility of the model stems from the ability to shrink it (to a very simple model considering atomic entity and relation embeddings only) or expand it (to a complex model by adding different contextual information) without affecting any other step in the learning procedure.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Representation Layer", "sec_num": "3.3" }, { "text": "Having observed a triplet (e s , r, e o ), we first use Eq. 7, 8 and 9 to compute entity and relation representations. We then use these embeddings to capture relational interaction between two entities using the following score function g(\u2022):", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Relational Score Function", "sec_num": "3.4" }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "g(e s , r, e o ) = \u03c3(z r T \u2022 (z e s z e o ))", "eq_num": "(10)" } ], "section": "Relational Score Function", "sec_num": "3.4" }, { "text": "where z r , z e s , z e o \u2208 R d are d-dimensional representations of entity and relationships as described below. \u03c3 is the nonlinear activation function and represent element-wise product.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Relational Score Function", "sec_num": "3.4" }, { "text": "4 Efficient Learning Procedure", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Relational Score Function", "sec_num": "3.4" }, { "text": "The complete parameter space of the model can be given by:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Objective Function", "sec_num": "4.1" }, { "text": "\u2126 = {{W i } 5 i=1 , W E , W R , W key , W val , W t , \u0398}.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Objective Function", "sec_num": "4.1" }, { "text": "To learn these parameters, we design a novel multitask objective function that jointly trains over two graphs. As identified earlier, the goal of our model is to leverage the available linkage information across graphs for optimizing the entity and relation embeddings such that they can explain the observed triplets across the graphs. Further, we want to leverage these optimized embeddings to match entities across graphs and expand the available linkage information. To achieve this goal, we define following two different loss functions catering to each learning task and jointly optimize over them as a multi-task objective to learn model parameters:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Objective Function", "sec_num": "4.1" }, { "text": "Relational Learning Loss. This is conventional loss function used to learn knowledge graph embeddings. Specifically, given a p-th triplet (e s , r, e o ) p from training set D, we sample C negative samples by replacing either head or tail entity and define a contrastive max margin function as shown in (Socher et al., 2013) :", "cite_spans": [ { "start": 303, "end": 324, "text": "(Socher et al., 2013)", "ref_id": "BIBREF36" } ], "ref_spans": [], "eq_spans": [], "section": "Objective Function", "sec_num": "4.1" }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "L rel = C c=1 max(0, \u03b3 \u2212 g(e s p , r p , e o p ) + g (e s c , r p , e o p ))", "eq_num": "(11)" } ], "section": "Objective Function", "sec_num": "4.1" }, { "text": "where, \u03b3 is margin, e s c represent corrupted entity and g (e s c , r p , e o p ) represent corrupted triplet score.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Objective Function", "sec_num": "4.1" }, { "text": "We design a novel loss function to leverage pairwise label set L. Given a triplet (e s X , r X , e o X ) from knowledge graph X, we first find the entity e + Y from graph Y that represent the same real-world entity as e s X . We then replace e s X with e + Y and compute score g(e + Y , r X , e o X ). Next, we find set of all entities E \u2212 Y from graph Y that has a negative label with entity e s X . We consider them analogous to the negative samples we generated for Eq. 11. We then propose the label learning loss function as:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Linkage Learning Loss:", "sec_num": null }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "L lab = Z z=1 max(0, \u03b3 \u2212 g(e + Y , r X , e o X ) + (g (e \u2212 Y , r X , e o X ) z ))", "eq_num": "(12)" } ], "section": "Linkage Learning Loss:", "sec_num": null }, { "text": "where, Z is the total number of negative labels for e X . \u03b3 is margin which is usually set to 1 and e \u2212 Y \u2208 E \u2212 Y represent entity from graph Y with which entity e s X had a negative label. Please note that this applies symmetrically for the triplets that originate from graph Y in the overall dataset. Note that if both entities of a triplet have labels, we will include both cases when computing the loss. Eq. 12 is inspired by Insight 1 and Insight 2 defined earlier in Section 2. Given a set D of N observed triplets across two graphs, we define complete multi-task objective as: where \u2126 is set of all model parameters and \u03bb is regularization hyper-parameter. b is weight hyperparameter used to attribute importance to each task. We train with mini-batch SGD procedure (Algorithm 1) using Adam Optimizer. Missing Positive Labels. It is expensive to obtain positive labels across multiple graphs and hence it is highly likely that many entities will not have positive labels available. For those entities, we will modify Eq. 12 to use the original triplet (e s X , r X , e o X ) in place of perturbed triplet g(e + Y , r X , e o X ) for the positive label. The rationale here again arises from Insight 2 wherein embeddings of two duplicate entities should be able to replace each other without affecting the score. Training Time Complexity. Most contextual information is pre-computed and available to all training steps which leads to constant time embedding lookup for those context. But for attribute network, embedding needs to be computed for each attribute separately and hence the complexity to compute score for one triplet is O(2a) where a is number of attributes. Also for training, we generate C negative samples for relational loss function and use Z negative labels for label loss function. Let k = C + Z. Hence, the training time complexity for a set of n triplets will be O(2ak * n) which is linear in number of triplets with a constant factor as ak << n for real world knowledge graphs. This is desirable as the number of triplets tend to be very large per graph in multi-relational settings. Memory Complexity. We borrow notations from (Nickel et al., 2016a) and describe the parameter complexity of our model in terms of the number of each component and corresponding embedding dimension requirements. Let", "cite_spans": [ { "start": 2156, "end": 2178, "text": "(Nickel et al., 2016a)", "ref_id": "BIBREF29" } ], "ref_spans": [], "eq_spans": [], "section": "Linkage Learning Loss:", "sec_num": null }, { "text": "EQUATION", "cite_spans": [], "ref_spans": [], "eq_spans": [ { "start": 0, "end": 8, "text": "EQUATION", "ref_id": "EQREF", "raw_str": "L(\u2126) = N i=1 [b\u2022L rel +(1\u2212b)\u2022L lab ]+\u03bb \u2126 2 2", "eq_num": "(13" } ], "section": "Linkage Learning Loss:", "sec_num": null }, { "text": "H a = 2 * N e H e +N r H r +N t H t +N k H k +N v H v .", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Linkage Learning Loss:", "sec_num": null }, { "text": "The parameter complexity of our model is:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Linkage Learning Loss:", "sec_num": null }, { "text": "H a * (H b + 1). Here, N e , N r , N t , N k , N v", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Linkage Learning Loss:", "sec_num": null }, { "text": "signify number of entities, relations, types, attribute keys and vocab size of attribute values across both datasets. Here H b is the output dimension of the hidden layer.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Linkage Learning Loss:", "sec_num": null }, { "text": "We evaluate LinkNBed and baselines on two real world knowledge graphs: D-IMDB (derived from large scale IMDB data snapshot) and D-FB (derived from large scale Freebase data snapshot). Table 5.1 provides statistics for our final dataset used in the experiments. Appendix B.1 provides complete details about dataset processing. ", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Datasets", "sec_num": "5.1" }, { "text": "We compare the performance of our method against state-of-the-art representation learning baselines that use neural embedding techniques to learn entity and relation representation. Specifically, we consider compositional methods of RESCAL (Nickel et al., 2011) as basic matrix factorization method, DISTMULT (Yang et al., 2015) as simple multiplicative model good for capturing symmetric relationships, and Complex (Trouillon et al., 2016) , an upgrade over DISTMULT that can capture asymmetric relationships using complex valued embeddings. We also compare against translational model of STransE that combined original structured embedding with TransE and has shown state-of-art performance in benchmark testing (Kadlec et al., 2017) . Finally, we compare with GAKE (Feng et al., 2016) , a model that captures context in entity and relationship representations. ", "cite_spans": [ { "start": 240, "end": 261, "text": "(Nickel et al., 2011)", "ref_id": "BIBREF31" }, { "start": 309, "end": 328, "text": "(Yang et al., 2015)", "ref_id": "BIBREF43" }, { "start": 416, "end": 440, "text": "(Trouillon et al., 2016)", "ref_id": "BIBREF40" }, { "start": 714, "end": 735, "text": "(Kadlec et al., 2017)", "ref_id": "BIBREF22" }, { "start": 768, "end": 787, "text": "(Feng et al., 2016)", "ref_id": "BIBREF16" } ], "ref_spans": [], "eq_spans": [], "section": "Baselines", "sec_num": "5.2" }, { "text": "We evaluate our model using two inference tasks: Link Prediction. Given a test triplet (e s , r, e o ), we first score this triplet using Eq. 10. We then replace e o with all other entities in the dataset and filter the resulting set of triplets as shown in (Bordes et al., 2013) . We score the remaining set of perturbed triplets using Eq. 10. All the scored triplets are sorted based on the scores and then the rank of the ground truth triplet is used for the evaluation. We use this ranking mechanism to compute HITS@10 (predicted rank \u2264 10) and reciprocal rank ( 1 rank ) of each test triplet. We report the mean over all test samples.", "cite_spans": [ { "start": 258, "end": 279, "text": "(Bordes et al., 2013)", "ref_id": "BIBREF3" } ], "ref_spans": [], "eq_spans": [], "section": "Evaluation Scheme", "sec_num": "5.3" }, { "text": "Entity Linkage. In alignment with Insight 2, we pose a novel evaluation scheme to perform entity linkage. Let there be two ground truth test sample triplets: (e X , e + Y , 1) representing a positive duplicate label and (e X , e \u2212 Y , 0) representing a negative duplicate label. Algorithm 2 outlines the procedure to compute linkage probability or score q (\u2208 [0, 1]) for the pair (e X , e Y ). We use L1 distance between the two vectors analogous Algorithm 2 Entity Linkage Score Computation Input: Test pair -(e X \u2208 X, e Y \u2208 Y ). Output: Linkage Score -q.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Evaluation Scheme", "sec_num": "5.3" }, { "text": "1. Collect all triplets involving e X from graph X and all triplets involving e Y from graph Y into a combined set O.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Evaluation Scheme", "sec_num": "5.3" }, { "text": "Let |O| = k. 2. Construct S orig \u2208 R k .", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Evaluation Scheme", "sec_num": "5.3" }, { "text": "For each triplet o \u2208 O, compute score g(o) using Eq. 10 and store the score in S orig . 3. Create triplet set O as following: if o \u2208 O contain e X \u2208 X then Replace e X with e Y to create perturbed triplet o and store it in O end if if o \u2208 O contain e Y \u2208 Y then Replace e Y with e X to create perturbed triplet o and store it in O end if 4. Construct S repl \u2208 R k . For each triplet o \u2208 O , compute score g(o ) using Eq. 10 and store the score in S repl . 5. Compute q. Elements in S orig and S repl have one-one correspondence so take the mean absolute difference:", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Evaluation Scheme", "sec_num": "5.3" }, { "text": "q = |S orig -S repl | 1 return q", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Evaluation Scheme", "sec_num": "5.3" }, { "text": "to Mean Absolute Error (MAE). In lieu of hard-labeling test pairs, we use score q to compute Area Under the Precision-Recall Curve (AUPRC).", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Evaluation Scheme", "sec_num": "5.3" }, { "text": "For the baselines and the unsupervised version (with no labels for entity linkage) of our model, we use second stage multilayer Neural Network as classifier for evaluating entity linkage. Appendix B.2 provides training configuration details.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Evaluation Scheme", "sec_num": "5.3" }, { "text": "Link Prediction Results. We train LinkNBed model jointly across two knowledge graphs and then perform inference over individual graphs to report link prediction reports. For baselines, we train each baseline on individual graphs and use parameters specific to the graph to perform link prediction inference over each individual graph. Table 5 .4 shows link prediction performance for all methods. Our model variant with attention mechanism outperforms all the baselines with 4.15% improvement over single graph state-of-the-art Complex model on D-IMDB and 8.23% improvement on D-FB dataset. D-FB is more challenging dataset to Hence closer performance of those two models aligns with expected outcome. We observed that the Neighborhood context alone provides only marginal improvements while the model benefits more from the use of attributes. Despite being marginal, attention mechanism also improves accuracy for both datasets. Compared to the baselines which are obtained by trained and evaluated on individual graphs, our superior performance demonstrates the effectiveness of multi-graph learning. Entity Linkage Results. We report entity linkage results for our method in two settings: a.) Supervised case where we train using both the objective functions. b.) Unsupervised case where we learn with only the relational loss function. The latter case resembles the baseline training where each model is trained separately on two graphs in an unsupervised manner. For performing the entity linkage in unsupervised case for all models, we first train a second stage of simple neural network classifier and then perform inference. In the supervised case, we use Algorithm 2 for performing the inference. Table 5 .4 demonstrates the performance of all methods on this task. Our method significantly outperforms all the baselines with 33.86% over second best baseline in supervised case and 17.35% better performance in unsupervised case. The difference in the performance of our method in two cases demonstrate that the two training objectives are helping one another by learning across the graphs. GAKE's superior performance on this task compared to the other state-of-the-art relational baselines shows the importance of using contex- Compositional Models learn representations by various composition operators on entity and relational embeddings. These models are multiplicative in nature and highly expressive but often suffer from scalability issues. Initial models include RESCAL (Nickel et al., 2011) that uses a relation specific weight matrix to explain triplets via pairwise interactions of latent features, Neural Tensor Network (Socher et al., 2013 ), more expressive model that combines a standard NN layer with a bilinear tensor layer and (Dong et al., 2014) that employs a concatenation-projection method to project entities and relations to lower dimensional space. Later, many sophisticated models (Neural Association Model , HoLE (Nickel et al., 2016b) ) have been proposed. Path based composition models (Toutanova et al., 2016) and contextual models GAKE (Feng et al., 2016) have been recently studied to capture more information from graphs. Recently, model like Complex (Trouillon et al., 2016) and Analogy (Liu et al., 2017) have demonstrated state-of-the art performance on relational learning tasks. Translational Models ( (Bordes et al., 2014) , (Bordes et al., 2011) , (Bordes et al., 2013) , (Wang et al., 2014) , (Lin et al., 2015) , (Xiao et al., 2016) ) learn representation by employing translational operators on the embeddings and optimizing based on their score. They offer an additive and efficient alternative to expensive multiplicative models. Due to their simplicity, they often loose expressive power. For a comprehensive survey of relational learning methods and empirical comparisons, we refer the readers to (Nickel et al., 2016a) , (Kadlec et al., 2017) , (Toutanova and Chen, 2015) and (Yang et al., 2015) . None of these methods address multi-graph relational learning and cannot be adapted to tasks like entity linkage in straightforward manner.", "cite_spans": [ { "start": 2488, "end": 2509, "text": "(Nickel et al., 2011)", "ref_id": "BIBREF31" }, { "start": 2642, "end": 2662, "text": "(Socher et al., 2013", "ref_id": "BIBREF36" }, { "start": 2755, "end": 2774, "text": "(Dong et al., 2014)", "ref_id": "BIBREF13" }, { "start": 2950, "end": 2972, "text": "(Nickel et al., 2016b)", "ref_id": "BIBREF30" }, { "start": 3025, "end": 3049, "text": "(Toutanova et al., 2016)", "ref_id": "BIBREF39" }, { "start": 3077, "end": 3096, "text": "(Feng et al., 2016)", "ref_id": "BIBREF16" }, { "start": 3194, "end": 3218, "text": "(Trouillon et al., 2016)", "ref_id": "BIBREF40" }, { "start": 3231, "end": 3249, "text": "(Liu et al., 2017)", "ref_id": "BIBREF26" }, { "start": 3348, "end": 3371, "text": "( (Bordes et al., 2014)", "ref_id": "BIBREF2" }, { "start": 3374, "end": 3395, "text": "(Bordes et al., 2011)", "ref_id": "BIBREF4" }, { "start": 3398, "end": 3419, "text": "(Bordes et al., 2013)", "ref_id": "BIBREF3" }, { "start": 3422, "end": 3441, "text": "(Wang et al., 2014)", "ref_id": "BIBREF41" }, { "start": 3444, "end": 3462, "text": "(Lin et al., 2015)", "ref_id": "BIBREF25" }, { "start": 3465, "end": 3484, "text": "(Xiao et al., 2016)", "ref_id": "BIBREF42" }, { "start": 3854, "end": 3876, "text": "(Nickel et al., 2016a)", "ref_id": "BIBREF29" }, { "start": 3879, "end": 3900, "text": "(Kadlec et al., 2017)", "ref_id": "BIBREF22" }, { "start": 3903, "end": 3929, "text": "(Toutanova and Chen, 2015)", "ref_id": "BIBREF38" }, { "start": 3934, "end": 3953, "text": "(Yang et al., 2015)", "ref_id": "BIBREF43" } ], "ref_spans": [ { "start": 335, "end": 342, "text": "Table 5", "ref_id": null }, { "start": 1706, "end": 1713, "text": "Table 5", "ref_id": null } ], "eq_spans": [], "section": "Predictive Analysis", "sec_num": "5.4" }, { "text": "Entity Resolution refers to resolving entities available in knowledge graphs with entity mentions in text. (Dredze et al., 2010) proposed entity disambiguation method for KB population, (He et al., 2013) learns entity embeddings for resolution, (Huang et al., 2015) propose a sophisticated DNN architecture for resolution, (Campbell et al., 2016) proposes entity resolution across multiple social domains, (Fang et al., 2016) jointly embeds text and knowledge graph to perform resolution while (Globerson et al., 2016) proposes Attention Mechanism for Collective Entity Resolution.", "cite_spans": [ { "start": 107, "end": 128, "text": "(Dredze et al., 2010)", "ref_id": "BIBREF14" }, { "start": 186, "end": 203, "text": "(He et al., 2013)", "ref_id": "BIBREF20" }, { "start": 245, "end": 265, "text": "(Huang et al., 2015)", "ref_id": "BIBREF21" }, { "start": 323, "end": 346, "text": "(Campbell et al., 2016)", "ref_id": "BIBREF6" }, { "start": 406, "end": 425, "text": "(Fang et al., 2016)", "ref_id": "BIBREF15" }, { "start": 494, "end": 518, "text": "(Globerson et al., 2016)", "ref_id": "BIBREF18" } ], "ref_spans": [], "eq_spans": [], "section": "Entity Resolution in Relational Data", "sec_num": "6.2" }, { "text": "Recently, learning over multiple graphs have gained traction. (Liu and Yang, 2016 ) divides a multi-relational graph into multiple homogeneous graphs and learns associations across them by employing product operator. Unlike our work, they do not learn across multiple multi-relational graphs. (Pujara and Getoor, 2016) provides logic based insights for cross learning, (Pershina et al., 2015) does pairwise entity matching across multirelational graphs and is very expensive, (Chen et al., 2017) learns embeddings to support multi-lingual learning and Big-Align (Koutra et al., 2013) tackles graph alignment problem efficiently for bipartite graphs. None of these methods learn latent representations or jointly train graph alignment and learning which is the goal of our work.", "cite_spans": [ { "start": 62, "end": 81, "text": "(Liu and Yang, 2016", "ref_id": "BIBREF27" }, { "start": 293, "end": 318, "text": "(Pujara and Getoor, 2016)", "ref_id": "BIBREF33" }, { "start": 369, "end": 392, "text": "(Pershina et al., 2015)", "ref_id": "BIBREF32" }, { "start": 476, "end": 495, "text": "(Chen et al., 2017)", "ref_id": "BIBREF10" }, { "start": 562, "end": 583, "text": "(Koutra et al., 2013)", "ref_id": "BIBREF23" } ], "ref_spans": [], "eq_spans": [], "section": "Learning across multiple graphs", "sec_num": "6.3" }, { "text": "We present a novel relational learning framework that learns entity and relationship embeddings across multiple graphs. The proposed representation learning framework leverage an efficient learning and inference procedure which takes into account the duplicate entities representing the same real-world entity in a multi-graph setting. We demonstrate superior accuracies on link prediction and entity linkage tasks compared to the existing approaches that are trained only on individual graphs. We believe that this work opens a new research direction in joint representation learning over multiple knowledge graphs. Many data driven organizations such as Google and Microsoft take the approach of constructing a unified super-graph by integrating data from multiple sources. Such unification has shown to significantly help in various applications, such as search, question answering, and personal assistance. To this end, there exists a rich body of work on linking entities and relations, and conflict resolution (e.g., knowledge fusion (Dong et al., 2014) . Still, the problem remains challenging for large scale knowledge graphs and this paper proposes a deep learning solution that can play a vital role in this construction process. In real-world setting, we envision our method to be integrated in a large scale system that would include various other components for tasks like conflict resolution, active learning and human-in-loop learning to ensure quality of constructed super-graph. However, we point out that our method is not restricted to such use cases-one can readily apply our method to directly make inference over multiple graphs to support applications like question answering and conversations.", "cite_spans": [ { "start": 1040, "end": 1059, "text": "(Dong et al., 2014)", "ref_id": "BIBREF13" } ], "ref_spans": [], "eq_spans": [], "section": "Concluding Remarks and Future Work", "sec_num": "7" }, { "text": "For future work, we would like to extend the current evaluation of our work from a two-graph setting to multiple graphs. A straightforward approach is to create a unified dataset out of more than two graphs by combining set of triplets as described in Section 2, and apply learning and inference on the unified graph without any major change in the methodology. Our inductive framework learns functions to encode contextual information and hence is graph independent. Alternatively, one can develop sophisticated approaches with iterative merging and learning over pairs of graphs until exhausting all graphs in an input collection.", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Concluding Remarks and Future Work", "sec_num": "7" } ], "back_matter": [ { "text": "We would like to give special thanks to Ben London, Tong Zhao, Arash Einolghozati, Andrew Borthwick and many others at Amazon for helpful comments and discussions. We thank the reviewers for their valuable comments and efforts towards improving our manuscript. This project was supported in part by NSF(IIS-1639792, IIS-1717916).", "cite_spans": [], "ref_spans": [], "eq_spans": [], "section": "Acknowledgments", "sec_num": null } ], "bib_entries": { "BIBREF0": { "ref_id": "b0", "title": "DBpedia: A nucleus for a web of open data", "authors": [ { "first": "S\u00f6ren", "middle": [], "last": "Auer", "suffix": "" }, { "first": "Christian", "middle": [], "last": "Bizer", "suffix": "" }, { "first": "Georgi", "middle": [], "last": "Kobilarov", "suffix": "" }, { "first": "Jens", "middle": [], "last": "Lehmann", "suffix": "" }, { "first": "Richard", "middle": [], "last": "Cyganiak", "suffix": "" }, { "first": "Zachary", "middle": [], "last": "Ives", "suffix": "" } ], "year": 2007, "venue": "The Semantic Web", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "S\u00f6ren Auer, Christian Bizer, Georgi Kobilarov, Jens Lehmann, Richard Cyganiak, and Zachary Ives. 2007. DBpedia: A nucleus for a web of open data. In The Semantic Web.", "links": null }, "BIBREF1": { "ref_id": "b1", "title": "Freebase: a collaboratively created graph database for structuring human knowledge", "authors": [ { "first": "Kurt", "middle": [], "last": "Bollacker", "suffix": "" }, { "first": "Colin", "middle": [], "last": "Evans", "suffix": "" }, { "first": "Praveen", "middle": [], "last": "Paritosh", "suffix": "" }, { "first": "Tim", "middle": [], "last": "Sturge", "suffix": "" }, { "first": "Jamie", "middle": [], "last": "Taylor", "suffix": "" } ], "year": 2008, "venue": "SIGMOD Conference", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Kurt Bollacker, Colin Evans, Praveen Paritosh, Tim Sturge, and Jamie Taylor. 2008. Freebase: a col- laboratively created graph database for structuring human knowledge. In SIGMOD Conference.", "links": null }, "BIBREF2": { "ref_id": "b2", "title": "A semantic matching energy function for learning with multi-relational data. Machine Learning", "authors": [ { "first": "Antoine", "middle": [], "last": "Bordes", "suffix": "" }, { "first": "Xavier", "middle": [], "last": "Glorot", "suffix": "" }, { "first": "Jason", "middle": [], "last": "Weston", "suffix": "" }, { "first": "Yoshua", "middle": [], "last": "Bengio", "suffix": "" } ], "year": 2014, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Antoine Bordes, Xavier Glorot, Jason Weston, and Yoshua Bengio. 2014. A semantic matching energy function for learning with multi-relational data. Ma- chine Learning.", "links": null }, "BIBREF3": { "ref_id": "b3", "title": "Translating embeddings for modeling multirelational data", "authors": [ { "first": "Antoine", "middle": [], "last": "Bordes", "suffix": "" }, { "first": "Nicolas", "middle": [], "last": "Usunier", "suffix": "" }, { "first": "Alberto", "middle": [], "last": "Garcia-Duran", "suffix": "" }, { "first": "Jason", "middle": [], "last": "Weston", "suffix": "" }, { "first": "Oksana", "middle": [], "last": "Yakhnenko", "suffix": "" } ], "year": 2013, "venue": "Advances in neural information processing systems", "volume": "", "issue": "", "pages": "2787--2795", "other_ids": {}, "num": null, "urls": [], "raw_text": "Antoine Bordes, Nicolas Usunier, Alberto Garcia- Duran, Jason Weston, and Oksana Yakhnenko. 2013. Translating embeddings for modeling multi- relational data. In Advances in neural information processing systems, pages 2787-2795.", "links": null }, "BIBREF4": { "ref_id": "b4", "title": "Learning structured embeddings of knowledge bases", "authors": [ { "first": "Antoine", "middle": [], "last": "Bordes", "suffix": "" }, { "first": "Jason", "middle": [], "last": "Weston", "suffix": "" }, { "first": "Ronan", "middle": [], "last": "Collobert", "suffix": "" }, { "first": "Yoshua", "middle": [], "last": "Bengio", "suffix": "" } ], "year": 2011, "venue": "AAAI", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Antoine Bordes, Jason Weston, Ronan Collobert, and Yoshua Bengio. 2011. Learning structured embed- dings of knowledge bases. In AAAI.", "links": null }, "BIBREF5": { "ref_id": "b5", "title": "Rdf graph alignment with bisimulation", "authors": [ { "first": "Peter", "middle": [], "last": "Buneman", "suffix": "" }, { "first": "Slawek", "middle": [], "last": "Staworko", "suffix": "" } ], "year": 2016, "venue": "Proc. VLDB Endow", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Peter Buneman and Slawek Staworko. 2016. Rdf graph alignment with bisimulation. Proc. VLDB Endow.", "links": null }, "BIBREF6": { "ref_id": "b6", "title": "Cross-domain entity resolution in social media", "authors": [ { "first": "W", "middle": [ "M" ], "last": "Campbell", "suffix": "" }, { "first": "Lin", "middle": [], "last": "Li", "suffix": "" }, { "first": "C", "middle": [], "last": "Dagli", "suffix": "" }, { "first": "J", "middle": [], "last": "Acevedo-Aviles", "suffix": "" }, { "first": "K", "middle": [], "last": "Geyer", "suffix": "" }, { "first": "J", "middle": [ "P" ], "last": "Campbell", "suffix": "" }, { "first": "C", "middle": [], "last": "Priebe", "suffix": "" } ], "year": 2016, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": { "arXiv": [ "arXiv:1608.01386v1" ] }, "num": null, "urls": [], "raw_text": "W. M. Campbell, Lin Li, C. Dagli, J. Acevedo- Aviles, K. Geyer, J. P. Campbell, and C. Priebe. 2016. Cross-domain entity resolution in social me- dia. arXiv:1608.01386v1.", "links": null }, "BIBREF8": { "ref_id": "b8", "title": "Toward an architecture for neverending language learning", "authors": [ { "first": "", "middle": [], "last": "Mitchell", "suffix": "" } ], "year": 2010, "venue": "Proceedings of the Twenty-Fourth Conference on Artificial Intelligence (AAAI 2010)", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Mitchell. 2010. Toward an architecture for never- ending language learning. In Proceedings of the Twenty-Fourth Conference on Artificial Intelligence (AAAI 2010).", "links": null }, "BIBREF9": { "ref_id": "b9", "title": "Personalized recommendations using knowledge graphs: A probabilistic logic programming approach", "authors": [ { "first": "Rose", "middle": [], "last": "Catherine", "suffix": "" }, { "first": "William", "middle": [], "last": "Cohen", "suffix": "" } ], "year": 2016, "venue": "Proceedings of the 10th ACM Conference on Recommender Systems", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Rose Catherine and William Cohen. 2016. Personal- ized recommendations using knowledge graphs: A probabilistic logic programming approach. In Pro- ceedings of the 10th ACM Conference on Recom- mender Systems.", "links": null }, "BIBREF10": { "ref_id": "b10", "title": "Multilingual knowledge graph embeddings for cross-lingual knowledge alignment", "authors": [ { "first": "Muhao", "middle": [], "last": "Chen", "suffix": "" }, { "first": "Yingtao", "middle": [], "last": "Tian", "suffix": "" }, { "first": "Mohan", "middle": [], "last": "Yang", "suffix": "" }, { "first": "Carlo", "middle": [], "last": "Zaniolo", "suffix": "" } ], "year": 2017, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Muhao Chen, Yingtao Tian, Mohan Yang, and Carlo Zaniolo. 2017. Multilingual knowledge graph em- beddings for cross-lingual knowledge alignment.", "links": null }, "BIBREF11": { "ref_id": "b11", "title": "Kbqa: Learning question answering over qa corpora and knowledge bases", "authors": [ { "first": "Wanyun", "middle": [], "last": "Cui", "suffix": "" }, { "first": "Yanghua", "middle": [], "last": "Xiao", "suffix": "" }, { "first": "Haixun", "middle": [], "last": "Wang", "suffix": "" }, { "first": "Yangqiu", "middle": [], "last": "Song", "suffix": "" }, { "first": "", "middle": [], "last": "Seung-Won", "suffix": "" }, { "first": "Wei", "middle": [], "last": "Hwang", "suffix": "" }, { "first": "", "middle": [], "last": "Wang", "suffix": "" } ], "year": 2017, "venue": "Proc. VLDB Endow", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Wanyun Cui, Yanghua Xiao, Haixun Wang, Yangqiu Song, Seung-won Hwang, and Wei Wang. 2017. Kbqa: Learning question answering over qa corpora and knowledge bases. Proc. VLDB Endow.", "links": null }, "BIBREF12": { "ref_id": "b12", "title": "Entity query feature expansion using knowledge base links", "authors": [ { "first": "Jeffrey", "middle": [], "last": "Dalton", "suffix": "" }, { "first": "Laura", "middle": [], "last": "Dietz", "suffix": "" }, { "first": "James", "middle": [], "last": "Allan", "suffix": "" } ], "year": 2014, "venue": "Proceedings of the 37th International ACM SIGIR Conference on Research & Development in Information Retrieval", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Jeffrey Dalton, Laura Dietz, and James Allan. 2014. Entity query feature expansion using knowledge base links. In Proceedings of the 37th International ACM SIGIR Conference on Research & Devel- opment in Information Retrieval.", "links": null }, "BIBREF13": { "ref_id": "b13", "title": "Knowledge vault: A web-scale approach to probabilistic knowledge fusion", "authors": [ { "first": "Xin", "middle": [], "last": "Dong", "suffix": "" }, { "first": "Evgeniy", "middle": [], "last": "Gabrilovich", "suffix": "" }, { "first": "Geremy", "middle": [], "last": "Heitz", "suffix": "" }, { "first": "Wilko", "middle": [], "last": "Horn", "suffix": "" }, { "first": "Ni", "middle": [], "last": "Lao", "suffix": "" }, { "first": "Kevin", "middle": [], "last": "Murphy", "suffix": "" }, { "first": "Thomas", "middle": [], "last": "Strohmann", "suffix": "" }, { "first": "Shaohua", "middle": [], "last": "Sun", "suffix": "" }, { "first": "Wei", "middle": [], "last": "Zhang", "suffix": "" } ], "year": 2014, "venue": "Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining", "volume": "", "issue": "", "pages": "601--610", "other_ids": {}, "num": null, "urls": [], "raw_text": "Xin Dong, Evgeniy Gabrilovich, Geremy Heitz, Wilko Horn, Ni Lao, Kevin Murphy, Thomas Strohmann, Shaohua Sun, and Wei Zhang. 2014. Knowledge vault: A web-scale approach to probabilistic knowl- edge fusion. In Proceedings of the 20th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 601-610.", "links": null }, "BIBREF14": { "ref_id": "b14", "title": "Entity disambiguation for knowledge base population", "authors": [ { "first": "Mark", "middle": [], "last": "Dredze", "suffix": "" }, { "first": "Paul", "middle": [], "last": "Mcnamee", "suffix": "" }, { "first": "Delip", "middle": [], "last": "Rao", "suffix": "" }, { "first": "Adam", "middle": [], "last": "Gerber", "suffix": "" }, { "first": "Tim", "middle": [], "last": "Finin", "suffix": "" } ], "year": 2010, "venue": "Proceedings of the 23rd International Conference on Computational Linguistics", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Mark Dredze, Paul McNamee, Delip Rao, Adam Ger- ber, and Tim Finin. 2010. Entity disambiguation for knowledge base population. In Proceedings of the 23rd International Conference on Computa- tional Linguistics.", "links": null }, "BIBREF15": { "ref_id": "b15", "title": "Entity disambiguation by knowledge and text jointly embedding", "authors": [ { "first": "Wei", "middle": [], "last": "Fang", "suffix": "" }, { "first": "Jianwen", "middle": [], "last": "Zhang", "suffix": "" }, { "first": "Dilin", "middle": [], "last": "Wang", "suffix": "" }, { "first": "Zheng", "middle": [], "last": "Chen", "suffix": "" }, { "first": "Ming", "middle": [], "last": "Li", "suffix": "" } ], "year": 2016, "venue": "CoNLL", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Wei Fang, Jianwen Zhang, Dilin Wang, Zheng Chen, and Ming Li. 2016. Entity disambiguation by knowl- edge and text jointly embedding. In CoNLL.", "links": null }, "BIBREF16": { "ref_id": "b16", "title": "Gake: Graph aware knowledge embedding", "authors": [ { "first": "Jun", "middle": [], "last": "Feng", "suffix": "" }, { "first": "Minlie", "middle": [], "last": "Huang", "suffix": "" }, { "first": "Yang", "middle": [], "last": "Yang", "suffix": "" }, { "first": "Xiaoyan", "middle": [], "last": "Zhu", "suffix": "" } ], "year": 2016, "venue": "COLING", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Jun Feng, Minlie Huang, Yang Yang, and Xiaoyan Zhu. 2016. Gake: Graph aware knowledge embedding. In COLING.", "links": null }, "BIBREF17": { "ref_id": "b17", "title": "Wikipedia-based semantic interpretation for natural language processing", "authors": [ { "first": "Evgeniy", "middle": [], "last": "Gabrilovich", "suffix": "" }, { "first": "Shaul", "middle": [], "last": "Markovitch", "suffix": "" } ], "year": 2009, "venue": "J. Artif. Int. Res", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Evgeniy Gabrilovich and Shaul Markovitch. 2009. Wikipedia-based semantic interpretation for natural language processing. J. Artif. Int. Res.", "links": null }, "BIBREF18": { "ref_id": "b18", "title": "Collective entity resolution with multi-focal attention", "authors": [ { "first": "Nevena", "middle": [], "last": "Amir Globerson", "suffix": "" }, { "first": "Soumen", "middle": [], "last": "Lazic", "suffix": "" }, { "first": "Amarnag", "middle": [], "last": "Chakrabarti", "suffix": "" }, { "first": "Michael", "middle": [], "last": "Subramanya", "suffix": "" }, { "first": "Fernando", "middle": [], "last": "Ringaard", "suffix": "" }, { "first": "", "middle": [], "last": "Pereira", "suffix": "" } ], "year": 2016, "venue": "Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Amir Globerson, Nevena Lazic, Soumen Chakrabarti, Amarnag Subramanya, Michael Ringaard, and Fer- nando Pereira. 2016. Collective entity resolution with multi-focal attention. In Proceedings of the 54th Annual Meeting of the Association for Compu- tational Linguistics.", "links": null }, "BIBREF19": { "ref_id": "b19", "title": "Representation learning on graphs: Methods and applications", "authors": [ { "first": "William", "middle": [ "L" ], "last": "Hamilton", "suffix": "" }, { "first": "Rex", "middle": [], "last": "Ying", "suffix": "" }, { "first": "Jure", "middle": [], "last": "Leskovec", "suffix": "" } ], "year": 2017, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": { "arXiv": [ "arXiv:1709.05584" ] }, "num": null, "urls": [], "raw_text": "William L. Hamilton, Rex Ying, and Jure Leskovec. 2017. Representation learning on graphs: Methods and applications. arXiv:1709.05584.", "links": null }, "BIBREF20": { "ref_id": "b20", "title": "Learning entity representation for entity disambiguation", "authors": [ { "first": "Zhengyan", "middle": [], "last": "He", "suffix": "" }, { "first": "Shujie", "middle": [], "last": "Liu", "suffix": "" }, { "first": "Mu", "middle": [], "last": "Li", "suffix": "" }, { "first": "Ming", "middle": [], "last": "Zhou", "suffix": "" }, { "first": "Longkai", "middle": [], "last": "Zhang", "suffix": "" }, { "first": "Houfeng", "middle": [], "last": "Wang", "suffix": "" } ], "year": 2013, "venue": "Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Zhengyan He, Shujie Liu, Mu Li, Ming Zhou, Longkai Zhang, and Houfeng Wang. 2013. Learning entity representation for entity disambiguation. In Pro- ceedings of the 51st Annual Meeting of the Associ- ation for Computational Linguistics.", "links": null }, "BIBREF21": { "ref_id": "b21", "title": "Leveraging deep neural networks and knowledge graphs for entity disambiguation", "authors": [ { "first": "Hongzhao", "middle": [], "last": "Huang", "suffix": "" }, { "first": "Larry", "middle": [], "last": "Heck", "suffix": "" }, { "first": "Heng", "middle": [], "last": "Ji", "suffix": "" } ], "year": 2015, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": { "arXiv": [ "arXiv:1504.07678v1" ] }, "num": null, "urls": [], "raw_text": "Hongzhao Huang, Larry Heck, and Heng Ji. 2015. Leveraging deep neural networks and knowledge graphs for entity disambiguation. arXiv:1504.07678v1.", "links": null }, "BIBREF22": { "ref_id": "b22", "title": "Knowledge base completion: Baselines strike back", "authors": [ { "first": "Rudolph", "middle": [], "last": "Kadlec", "suffix": "" }, { "first": "Ondrej", "middle": [], "last": "Bajgar", "suffix": "" } ], "year": 2017, "venue": "Proceedings of the 2nd Workshop on Representation Learning for NLP", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Rudolph Kadlec, Ondrej Bajgar, and Jan Kleindienst. 2017. Knowledge base completion: Baselines strike back. In Proceedings of the 2nd Workshop on Rep- resentation Learning for NLP.", "links": null }, "BIBREF23": { "ref_id": "b23", "title": "Big-align: Fast bipartite graph alignment", "authors": [ { "first": "Danai", "middle": [], "last": "Koutra", "suffix": "" }, { "first": "Hanghang", "middle": [], "last": "Tong", "suffix": "" }, { "first": "David", "middle": [], "last": "Lubensky", "suffix": "" } ], "year": 2013, "venue": "2013 IEEE 13th International Conference on Data Mining", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Danai Koutra, HangHang Tong, and David Lubensky. 2013. Big-align: Fast bipartite graph alignment. In 2013 IEEE 13th International Conference on Data Mining.", "links": null }, "BIBREF24": { "ref_id": "b24", "title": "Distributed representations of sentences and documents", "authors": [ { "first": "Quoc", "middle": [], "last": "Le", "suffix": "" }, { "first": "Tomas", "middle": [], "last": "Mikolov", "suffix": "" } ], "year": 2014, "venue": "Proceedings of the 31st International Conference on Machine Learning", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Quoc Le and Tomas Mikolov. 2014. Distributed repre- sentations of sentences and documents. In Proceed- ings of the 31st International Conference on Ma- chine Learning.", "links": null }, "BIBREF25": { "ref_id": "b25", "title": "Learning entity and relation embeddings for knowledge graph completion", "authors": [ { "first": "Yankai", "middle": [], "last": "Lin", "suffix": "" }, { "first": "Zhiyuan", "middle": [], "last": "Liu", "suffix": "" }, { "first": "Maosong", "middle": [], "last": "Sun", "suffix": "" }, { "first": "Xuan", "middle": [], "last": "Zhu", "suffix": "" } ], "year": 2015, "venue": "AAAI Conference on Artificial Intelligence", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Yankai Lin, Zhiyuan Liu, Maosong Sun, and Xuan Zhu. 2015. Learning entity and relation embeddings for knowledge graph completion. AAAI Conference on Artificial Intelligence.", "links": null }, "BIBREF26": { "ref_id": "b26", "title": "Analogical inference for multi-relatinal embeddings", "authors": [ { "first": "Hanxiao", "middle": [], "last": "Liu", "suffix": "" }, { "first": "Yuexin", "middle": [], "last": "Wu", "suffix": "" }, { "first": "Yimin", "middle": [], "last": "Yang", "suffix": "" } ], "year": 2017, "venue": "Proceedings of the 34th International Conference on Machine Learning", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Hanxiao Liu, Yuexin Wu, and Yimin Yang. 2017. Ana- logical inference for multi-relatinal embeddings. In Proceedings of the 34th International Conference on Machine Learning.", "links": null }, "BIBREF27": { "ref_id": "b27", "title": "Cross-graph learning of multi-relational associations", "authors": [ { "first": "Hanxiao", "middle": [], "last": "Liu", "suffix": "" }, { "first": "Yimin", "middle": [], "last": "Yang", "suffix": "" } ], "year": 2016, "venue": "Proceedings of the 33rd International Conference on Machine Learning", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Hanxiao Liu and Yimin Yang. 2016. Cross-graph learn- ing of multi-relational associations. In Proceedings of the 33rd International Conference on Machine Learning.", "links": null }, "BIBREF28": { "ref_id": "b28", "title": "Probabilistic reasoning via deep learning: Neural association models", "authors": [ { "first": "Quan", "middle": [], "last": "Liu", "suffix": "" }, { "first": "Hui", "middle": [], "last": "Jiang", "suffix": "" }, { "first": "Andrew", "middle": [], "last": "Evdokimov", "suffix": "" }, { "first": "Zhen-Hua", "middle": [], "last": "Ling", "suffix": "" }, { "first": "Xiaodan", "middle": [], "last": "Zhu", "suffix": "" }, { "first": "Si", "middle": [], "last": "Wei", "suffix": "" }, { "first": "Yu", "middle": [], "last": "Hu", "suffix": "" } ], "year": 2016, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": { "arXiv": [ "arXiv:1603.07704v2" ] }, "num": null, "urls": [], "raw_text": "Quan Liu, Hui Jiang, Andrew Evdokimov, Zhen-Hua Ling, Xiaodan Zhu, Si Wei, and Yu Hu. 2016. Prob- abilistic reasoning via deep learning: Neural associ- ation models. arXiv:1603.07704v2.", "links": null }, "BIBREF29": { "ref_id": "b29", "title": "A review of relational machine learning for knowledge graphs", "authors": [ { "first": "Maximilian", "middle": [], "last": "Nickel", "suffix": "" }, { "first": "Kevin", "middle": [], "last": "Murphy", "suffix": "" }, { "first": "Volker", "middle": [], "last": "Tresp", "suffix": "" }, { "first": "Evgeniy", "middle": [], "last": "Gabrilovich", "suffix": "" } ], "year": 2016, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Maximilian Nickel, Kevin Murphy, Volker Tresp, and Evgeniy Gabrilovich. 2016a. A review of relational machine learning for knowledge graphs. Proceed- ings of the IEEE.", "links": null }, "BIBREF30": { "ref_id": "b30", "title": "Holographic embeddings of knowledge graphs", "authors": [ { "first": "Maximilian", "middle": [], "last": "Nickel", "suffix": "" }, { "first": "Lorenzo", "middle": [], "last": "Rosasco", "suffix": "" }, { "first": "Tomaso", "middle": [], "last": "Poggio", "suffix": "" } ], "year": 2016, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Maximilian Nickel, Lorenzo Rosasco, and Tomaso Poggio. 2016b. Holographic embeddings of knowl- edge graphs.", "links": null }, "BIBREF31": { "ref_id": "b31", "title": "A three-way model for collective learning on multi-relational data", "authors": [ { "first": "Maximilian", "middle": [], "last": "Nickel", "suffix": "" }, { "first": "Hans-Peter", "middle": [], "last": "Volker Tresp", "suffix": "" }, { "first": "", "middle": [], "last": "Kriegel", "suffix": "" } ], "year": 2011, "venue": "Proceedings of the 28th International Conference on Machine Learning (ICML-11)", "volume": "", "issue": "", "pages": "809--816", "other_ids": {}, "num": null, "urls": [], "raw_text": "Maximilian Nickel, Volker Tresp, and Hans-Peter Kriegel. 2011. A three-way model for collective learning on multi-relational data. In Proceedings of the 28th International Conference on Machine Learning (ICML-11), pages 809-816.", "links": null }, "BIBREF32": { "ref_id": "b32", "title": "Holistic entity matching across knowledge graphs", "authors": [ { "first": "Maria", "middle": [], "last": "Pershina", "suffix": "" }, { "first": "Mohamed", "middle": [], "last": "Yakout", "suffix": "" }, { "first": "Kaushik", "middle": [], "last": "Chakrabarti", "suffix": "" } ], "year": 2015, "venue": "2015 IEEE International Conference on Big Data (Big Data)", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Maria Pershina, Mohamed Yakout, and Kaushik Chakrabarti. 2015. Holistic entity matching across knowledge graphs. In 2015 IEEE International Con- ference on Big Data (Big Data).", "links": null }, "BIBREF33": { "ref_id": "b33", "title": "Generic statistical relational entity resolution in knowledge graphs", "authors": [ { "first": "Jay", "middle": [], "last": "Pujara", "suffix": "" }, { "first": "Lise", "middle": [], "last": "Getoor", "suffix": "" } ], "year": 2016, "venue": "Sixth International Workshop on Statistical Relational AI", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Jay Pujara and Lise Getoor. 2016. Generic statisti- cal relational entity resolution in knowledge graphs. In Sixth International Workshop on Statistical Rela- tional AI.", "links": null }, "BIBREF35": { "ref_id": "b35", "title": "Deep feature learning for graphs", "authors": [ { "first": "Ahmed", "middle": [], "last": "", "suffix": "" } ], "year": 2017, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": { "arXiv": [ "arXiv:1704.08829" ] }, "num": null, "urls": [], "raw_text": "Ahmed. 2017. Deep feature learning for graphs. arXiv:1704.08829.", "links": null }, "BIBREF36": { "ref_id": "b36", "title": "Reasoning with neural tensor networks for knowledge base completion", "authors": [ { "first": "Richard", "middle": [], "last": "Socher", "suffix": "" }, { "first": "Danqi", "middle": [], "last": "Chen", "suffix": "" }, { "first": "D", "middle": [], "last": "Christopher", "suffix": "" }, { "first": "Andrew", "middle": [], "last": "Manning", "suffix": "" }, { "first": "", "middle": [], "last": "Ng", "suffix": "" } ], "year": 2013, "venue": "Advances in Neural Information Processing Systems", "volume": "", "issue": "", "pages": "926--934", "other_ids": {}, "num": null, "urls": [], "raw_text": "Richard Socher, Danqi Chen, Christopher D Manning, and Andrew Ng. 2013. Reasoning with neural ten- sor networks for knowledge base completion. In Advances in Neural Information Processing Systems, pages 926-934.", "links": null }, "BIBREF37": { "ref_id": "b37", "title": "Yago: A core of semantic knowledge", "authors": [ { "first": "Fabian", "middle": [ "M" ], "last": "Suchanek", "suffix": "" }, { "first": "Gjergji", "middle": [], "last": "Kasneci", "suffix": "" }, { "first": "Gerhard", "middle": [], "last": "Weikum", "suffix": "" } ], "year": 2007, "venue": "Proceedings of the 16th International Conference on World Wide Web", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Fabian M. Suchanek, Gjergji Kasneci, and Gerhard Weikum. 2007. Yago: A core of semantic knowl- edge. In Proceedings of the 16th International Con- ference on World Wide Web.", "links": null }, "BIBREF38": { "ref_id": "b38", "title": "Observed versus latent features for knowledge base and text inference", "authors": [ { "first": "Kristina", "middle": [], "last": "Toutanova", "suffix": "" }, { "first": "Danqi", "middle": [], "last": "Chen", "suffix": "" } ], "year": 2015, "venue": "ACL", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Kristina Toutanova and Danqi Chen. 2015. Observed versus latent features for knowledge base and text inference. In ACL.", "links": null }, "BIBREF39": { "ref_id": "b39", "title": "Compositional learning of embeddings for relation paths in knowledge bases and text", "authors": [ { "first": "Kristina", "middle": [], "last": "Toutanova", "suffix": "" }, { "first": "Victoria", "middle": [], "last": "Xi", "suffix": "" }, { "first": "Wen-Tau", "middle": [], "last": "Lin", "suffix": "" }, { "first": "Hoifung", "middle": [], "last": "Yih", "suffix": "" }, { "first": "Chris", "middle": [], "last": "Poon", "suffix": "" }, { "first": "", "middle": [], "last": "Quirk", "suffix": "" } ], "year": 2016, "venue": "Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics", "volume": "1", "issue": "", "pages": "1434--1444", "other_ids": {}, "num": null, "urls": [], "raw_text": "Kristina Toutanova, Xi Victoria Lin, Wen-tau Yih, Hoi- fung Poon, and Chris Quirk. 2016. Compositional learning of embeddings for relation paths in knowl- edge bases and text. In Proceedings of the 54th An- nual Meeting of the Association for Computational Linguistics, volume 1, pages 1434-1444.", "links": null }, "BIBREF40": { "ref_id": "b40", "title": "Complex embeddings for simple link prediction", "authors": [ { "first": "Theo", "middle": [], "last": "Trouillon", "suffix": "" }, { "first": "Johannes", "middle": [], "last": "Welbl", "suffix": "" }, { "first": "Sebastian", "middle": [], "last": "Riedel", "suffix": "" }, { "first": "Eric", "middle": [], "last": "Gaussier", "suffix": "" }, { "first": "Guillaume", "middle": [], "last": "Bouchard", "suffix": "" } ], "year": 2016, "venue": "Proceedings of the 33rd International Conference on Machine Learning", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Theo Trouillon, Johannes Welbl, Sebastian Riedel, Eric Gaussier, and Guillaume Bouchard. 2016. Complex embeddings for simple link prediction. In Proceed- ings of the 33rd International Conference on Ma- chine Learning.", "links": null }, "BIBREF41": { "ref_id": "b41", "title": "Knowledge graph embedding by translating on hyperplanes", "authors": [ { "first": "Zhen", "middle": [], "last": "Wang", "suffix": "" }, { "first": "Jianwen", "middle": [], "last": "Zhang", "suffix": "" }, { "first": "Jianlin", "middle": [], "last": "Feng", "suffix": "" }, { "first": "Zheng", "middle": [], "last": "Chen", "suffix": "" } ], "year": 2014, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Zhen Wang, Jianwen Zhang, Jianlin Feng, and Zheng Chen. 2014. Knowledge graph embedding by trans- lating on hyperplanes.", "links": null }, "BIBREF42": { "ref_id": "b42", "title": "Transg: A generative model for knowledge graph embedding", "authors": [ { "first": "Han", "middle": [], "last": "Xiao", "suffix": "" }, { "first": "Minlie", "middle": [], "last": "Huang", "suffix": "" }, { "first": "Xiaoyan", "middle": [], "last": "Zhu", "suffix": "" } ], "year": 2016, "venue": "Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics", "volume": "", "issue": "", "pages": "", "other_ids": {}, "num": null, "urls": [], "raw_text": "Han Xiao, Minlie Huang, and Xiaoyan Zhu. 2016. Transg: A generative model for knowledge graph embedding. In Proceedings of the 54th Annual Meeting of the Association for Computational Lin- guistics.", "links": null }, "BIBREF43": { "ref_id": "b43", "title": "Embedding entities and relations for learning and inference in knowledge bases", "authors": [ { "first": "Bishan", "middle": [], "last": "Yang", "suffix": "" }, { "first": "Wen-Tau", "middle": [], "last": "Yih", "suffix": "" }, { "first": "Xiaodong", "middle": [], "last": "He", "suffix": "" }, { "first": "Jianfeng", "middle": [], "last": "Gao", "suffix": "" }, { "first": "Li", "middle": [], "last": "Deng", "suffix": "" } ], "year": 2015, "venue": "", "volume": "", "issue": "", "pages": "", "other_ids": { "arXiv": [ "arXiv:1412.6575" ] }, "num": null, "urls": [], "raw_text": "Bishan Yang, Wen-tau Yih, Xiaodong He, Jianfeng Gao, and Li Deng. 2015. Embedding entities and relations for learning and inference in knowledge bases. arXiv:1412.6575.", "links": null } }, "ref_entries": { "FIGREF0": { "type_str": "figure", "num": null, "text": "LinkNBed Architecture Overview -one step score computation for a given triplet (e s , r, e o ). The Attribute embeddings are not simple lookups but they are learned as shown in Eq 3", "uris": null }, "TABREF0": { "type_str": "table", "text": "LinkNBed mini-batch Training Input: Mini-batch M, Negative Sample Size C, Negative Label Size Z, Attribute data att data, Neighborhood data nhbr data, Type data type data, Positive Label Dict pos dict, Negative Label Dict neg dict Output: Mini-batch Loss L M .", "html": null, "num": null, "content": "
Algorithm 1 L M = 0
score pos = []; score neg = []; score pos lab =
[]; score neg lab = []
for i = 0 to size(M) do
input tuple = M[i] = (e s , r, e o )
sc = compute triplet score(e s , r, e o ) (Eq. 10)
score pos.append(sc)
for j = 0 to C do
Select e s c from entity list such that e s c = e s and e s c = e o and (e s \u2208 D c , r, e o ) / sc neg = compute triplet score(e s c , r, e o )
score neg.append(sc neg)
end for
if e l)
end if
for k = 0 to Z do
Select e s\u2212 from neg dict
sc neg l = compute triplet score(e s\u2212 , r, e o )
score neg lab.append(sc neg l)
end for
end for
L M += compute minibatch loss(score pos,
score neg,score pos lab,score neg lab)
(Eq. 13)
Back-propagate errors and update parameters \u2126
return L M
)
" }, "TABREF2": { "type_str": "table", "text": "Statistics for Datasets: D-IMDB and D-FB", "html": null, "num": null, "content": "" }, "TABREF5": { "type_str": "table", "text": "", "html": null, "num": null, "content": "
: Link Prediction Results on both datasets
learn as it has a large set of sparse relationships,
types and attributes and it has an order of magni-
tude lesser relational evidence (number of triplets)
compared to D-IMDB. Hence, LinkNBed's pro-
nounced improvement on D-FB demonstrates the
effectiveness of the model. The simplest version of
LinkNBed with only entity embeddings resembles
DISTMULT model with different objective func-
tion.
" }, "TABREF7": { "type_str": "table", "text": "", "html": null, "num": null, "content": "
: Entity Linkage Results -Unsupervised
case uses classifier at second step
tual information for entity linkage. Performance of
other variants of our model again demonstrate that
attribute information is more helpful than neigh-
borhood context and attention provides marginal
improvements. We provide further insights with
examples and detailed discussion on entity linkage
task in Appendix A.
6 Related Work
6.1 Neural Embedding Methods for
Relational Learning
" } } } }