Leveraging Distributed Representations of Elements in Triples for Predicate Linking
Knowledge graphs (KGs) play a crucial role in many modern applications. Many open information extraction approaches propose the extraction of triples from natural language text in order to populate knowledge. Nonetheless, most approaches do not consider forming links between the extracted triples and the KG triples, especially for predicates. Predicate linking is used to identify the predicate in a KG that exactly corresponds to an extracted predicate; this allows the avoidance of the heterogeneous problem. Resolving the heterogeneous problem can increase searchability over KGs. Although there have been a few studies that considered linking predicates, most of them have relied on statistical knowledge patterns, which are not able to generate all possible patterns. In this paper, we introduce distributed representations of elements in triples and leverage them for computing the similarity between predicates in order to find links that would not appear in statistical patterns. In the experiment, the results show that leveraging the distributed representations of triple elements can discover links between identical predicates, which cannot be achieved by the statistical pattern approach. As a result, our approach outperformed the traditional baseline for the predicate linking task.
KeywordsKnowledge graph Predicate linking Knowledge pattern
This work was partially supported by NEDO (New Energy and Industrial Technology Development Organization).
- 3.Carlson, A., Betteridge, J., Kisiel, B., Settles, B., Hruschka Jr., E.R., Mitchell, T.M.: Toward an architecture for never-ending language learning. In: Proceedings of AAAI (2010)Google Scholar
- 5.Exner, P., Nugues, P.: Entity extraction: from unstructured text to DBpedia RDF triples. In: The Web of Linked Entities Workshop, pp. 58–69. CEUR-WS (2012)Google Scholar
- 6.Fader, A., Soderland, S., Etzioni, O.: Identifying relations for open information extraction. In: Proceedings of the Conference on Empirical Methods in Natural Language Processing, pp. 1535–1545. ACL (2011)Google Scholar
- 8.Lee, H., Peirsman, Y., Chang, A., Chambers, N., Surdeanu, M., Jurafsky, D.: Stanford’s multi-pass sieve coreference resolution system at the CoNLL-2011 shared task. In: Proceedings of the 15th Conference on Computational Natural Language Learning: Shared Task, pp. 28–34. ACL (2011)Google Scholar
- 9.Mendes, P.N., Jakob, M., García-Silva, A., Bizer, C.: DBpedia spotlight: shedding light on the web of documents. In: Proceedings of the 7th International Conference on Semantic Systems, pp. 1–8. ACM (2011)Google Scholar
- 10.Mikolov, T., Sutskever, I., Chen, K., Corrado, G.S., Dean, J.: Distributed representations of words and phrases and their compositionality. In: Proceedings of Advances in Neural Information Processing Systems, pp. 3111–3119 (2013)Google Scholar
- 11.Schmitz, M., Bart, R., Soderland, S., Etzioni, O.: Open language learning for information extraction. In: Proceedings of the Joint Conference on EMNLP, pp. 523–534. ACL (2012)Google Scholar
- 12.Zhang, Z., Gentile, A.L., Blomqvist, E., Augenstein, I., Ciravegna, F.: Statistical knowledge patterns: identifying synonymous relations in large linked datasets. In: Alani, H., et al. (eds.) ISWC 2013. LNCS, vol. 8218, pp. 703–719. Springer, Heidelberg (2013). doi: 10.1007/978-3-642-41335-3_44 CrossRefGoogle Scholar