Open Knowledge Graphs Canonicalization using Variational Autoencoders

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, pages 10379–10394 November 7–11, 2021. c⃝2021 Association for Computational Linguistics 10379 Open Knowledge Graphs Canonicalization using Variational Autoencoders Sarthak Dash, Gaetano Rossiello, Sugato Bagchi, Nandana Mihindukulasooriya, Alfio Gliozzo IBM Research AI, Thomas J. Watson Research Center Yorktown Heights, NY Abstract Noun phrases and Relation phrases in open knowledge graphs are not canonicalized, lead- ing to an explosion of redundant and ambigu- ous subject-relation-object triples. Existing ap- proaches to solve this problem take a two-step approach. First, they generate embedding rep- resentations for both noun and relation phrases, then a clustering algorithm is used to group them using the embeddings as features. In this work, we propose Canonicalizing Using Vari- ational Autoencoders (CUVA)1, a joint model to learn both embeddings and cluster assign- ments in an end-to-end approach, which leads to a better vector representation for the noun and relation phrases. Our evaluation over mul- tiple benchmarks shows that CUVA outper- forms the existing state-of-the-art approaches. Moreover, we introduce CANONICNELL, a novel dataset to evaluate entity canonicaliza- tion systems. 1 Introduction Open Information Extraction (OpenIE) methods (Fader et al., 2011a; Stanovsky et al., 2018) can be used to extract triples in the form (noun phrase, relation phrase, noun phrase) from given text cor- pora in an unsupervised way without requiring a pre-defined ontology schema. This makes them suitable to build large Open Knowledge Graphs (OpenKGs) from huge collections of unstructured text documents, thereby making the usage of Ope- nIE methods highly adaptable to new domains. Although OpenIE methods are highly adaptable, one major shortcoming of OpenKGs is that Noun Phrases (NPs) and Relation Phrases (RPs) are not canonicalized. This means that two NPs (or RPs) having different surface forms, but referring to the same entity (or relation) in a canonical KB, are treated differently. Consider the following triples 1https://github.com/IBM/ Open-KG-canonicalization as an example: (NBC-TV, has headquarters in, NYC), (NBC Television, is in, New York City) and (NBC-TV, has main office in, NYC). Looking at the previous example, both OpenIE methods and associated Open KGs would not have any knowl- edge that NYC and New York City refer to the same entity, or has headquarters in and has main office in are similar relations. Moreover, while it is true that similar relations will have same argument types (see the previous example), the converse need not hold true. For example, given the following two triples (X, is born in, Y) and (X, has died in, Y) in an Open KG, where X is of type Person and Y is of type Location, does not imply is born in and has died in are similar relations. Thus, the task of canonicalizing NPs and RPs within an Open KG is significant. Otherwise, Open KGs will have an explosion of redundant facts, which is highly undesirable, for the following rea- sons. Firstly, redundant facts use a higher mem- ory footprint. Secondly, querying an Open KG is likely to yield sub-optimal results, for e.g. it …

Content truncated to 500 words