Typed Tensor Decomposition of Knowledge Bases for Relation Extraction

By Kai-Wei Chang, Wen-tau Yih, Bishan Yang, Christopher Meek. 2014

208 citations recorded by the source corpus.

While relation extraction has traditionally been viewed as a task relying solely on textual data, recent work has shown that by taking as input existing facts in the form of entity-relation triples from both knowledge bases and textual data, the performance of relation extraction can be improved significantly. Following this new paradigm, we propose a tensor decomposition approach for knowledge base embedding that is highly scalable, and is especially suitable for relation extraction. By leveraging relational domain knowledge about entity type information, our learning algorithm is significantly faster than previous approaches and is better able to discover new relations missing from the database. In addition, when applied to a relation extraction task, our approach alone is comparable to several existing systems, and improves the weighted mean average precision of a state-of-theart method by 10 points when used as a subcomponent.

Subfields: Artificial Intelligence, Computational Mathematics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computational Mathematics

Citations in the corpus

Modeling Relational Data with Graph Convolutional Networks (2017, methodology): “We expect this to be a promising avenue for future research.”
Embedding Entities and Relations for Learning and Inference in Knowledge Bases (2014, methodology): “Inspired by (Chang et al., 2014), we design a new evaluation setting where the predicted entities are automatically ﬁltered according to “entity types” (entities that appear as the subjects/objects of a relation have the same type deﬁned by that relation).”
Representing Text for Joint Embedding of Text and Knowledge Bases (2015, methodology): “We also limit the candidate entities to ones that have types consistent with the position in the relation triple (Chang et al., 2014; Yang et al., 2015), where the types are approximated following Toutanova and Chen (2015).”
Observed versus latent features for knowledge base and text inference (2015, background): “In some settings we also limit the candidate entities to ones that have types consistent with the position in the relation triple (Chang et al., 2014; Yang et al., 2015).”
Semantically Smooth Knowledge Graph Embedding (2015, methodology): “In the second group, KGs are represented as tensors, and embedding is performed via tensor factorization or collective matrix factorization techniques (Singh and Gordon, 2008; Nickel et al., 2011; Chang et al., 2014).”
Learning Entity Type Embeddings for Knowledge Graph Completion (2017, background): “Entity type information in KGs is widely used in various NLP tasks such as relation extraction [4], entity linking [6], and question answering [1, 16].”
Tensors for Data Mining and Data Fusion (2016, methodology): “Following the examples of Nickel et al. [2012] and Chang et al. [2014] use a RESCAL-inspired decomposition to model and analyze knowledge base data.”
CORE: Context-Aware Open Relation Extraction with Factorization Machines (2015, methodology): “Chang et al. (2014) have also shown”
Relation Extraction Using Distant Supervision (2018, background): “(4) The combined model (NFE) defines θ r , t The authors of [10] extended RESCAL , a tensor decomposition approach described in [44], by The data is encoded as a tensor X ∈ { 0 , 1 } n × n × m , where n is the number of entities, m is the number of relations, X i , j , k = 1 if and only if a fact r…”
Discrete-State Variational Autoencoders for Joint Discovery and Factorization of Relations (2016, methodology): “Knowledge base completion via matrix or tensor factorization has received a lot of attention in the past years (Bordes et al., 2011; Jenatton et al., 2012; Weston et al., 2013; Bordes et al., 2013; Socher et al., 2013; Garcı́aDurán et al., 2014; Bordes et al., 2014; Lin et al., 2015b; Chang et al., ”
Lifted Rule Injection for Relation Embeddings (2016, background): “Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing, pages 1389–1399, Austin, Texas, November 1-5, 2016. c©2016 Association for Computational Linguistics”
Context-Dependent Knowledge Graph Embedding (2015, methodology): “For other KG embedding models, please refer to (Nickel et al., 2011; Riedel et al., 2013; Wang et al., 2014; Chang et al., 2014).”
Knowledge Graph Completion via Complex Tensor Factorization (2017, methodology): “Chang et al. (2014) enhance the RESCAL model to take into account information about the entity types.”
Compositional Learning of Embeddings for Relation Paths in Knowledge Base and Text (2016, background): “Originally proposed as an alternative statistical relational learning method, the knowledge base embedding approach has gained a signiﬁ-cant amount of attention, due to its simple prediction time computation and strong empirical performance (Nickel et al., 2011; Chang et al., 2014).”
Leveraging Knowledge Bases in LSTMs for Improving Machine Reading (2017, background): “, 2015), relation extraction (Weston et al., 2013; Chang et al., 2014; Riedel et al., 2013), and question answering (Miller et al.”