Abstract: As a joint-optimization problem which simultaneously fulfills two different but correlated embedding tasks (i.e., entity embedding and relation embedding), knowledge embedding problem is solved in a joint embedding scheme. In this embedding scheme, we design a joint compatibility scoring function to quantitatively evaluate the relational facts with respect to entities and relations, and further incorporate the scoring function into the max-margin structure learning process that explicitly learns the embedding vectors of entities and relations using the context information of the knowledge base. By optimizing the joint problem, our design is capable of effectively capturing the intrinsic topological structures in the learned embedding spaces. Experimental results demonstrate the effectiveness of our embedding scheme in characterizing the semantic correlations among different relation units, and in relation prediction for knowledge inference.

基于代价敏感学习的实体–关系联合知识嵌入

[1]Bollacker, K., Evans, C., Paritosh, P., et al., 2008. Freebase: a collaboratively created graph database for structuring human knowledge. Proc. Int. Conf. on ACM SIGMOD Management of Data, p.1247-1250.

[2]Bordes, A., Weston, J., Collobert, R., et al., 2011. Learning structured embeddings of knowledge bases. Proc. Comput. Sci., 108:345-354.

[3]Bordes, A., Usunier, N., Garcia-Duran, A., et al., 2013. Translating embeddings for modeling multi-relational data. NIPS, p.2787-2795.

[4]Bordes, A., Glorot, X., Weston, J., et al., 2014. A semantic matching energy function for learning with multi-relational data. Mach. Learn., 94(2):233-259.

[5]Chang, K.W., Yih, W.T., Meek, C., 2013. Multi-relational latent semantic analysis. Proc. Conf. on Empirical Methods in Natural Language Processing, p.1602-1612.

[6]Duchi, J., Hazan, E., Singer, Y., 2011. Adaptive subgradient methods for online learning and stochastic optimization. J. Mach. Learn. Res., 12:2121-2159.

[7]Elisseeff, A., Weston, J., 2001. A kernel method for multi-labelled classification. NIPS, p.681-687.

[8]García-Durán, A., Bordes, A., Usunier, N., 2015. Composing Relationships with Translations. Technical Report linebreak No. hal-01167811, CNRS-Heudiasyc, Compiègne.

[9]Getoor, L., Mihalkova, L., 2011. Learning statistical models from relational data. Proc. Int. Conf. on ACM SIGMOD Management of Data, p.1195-1198.

[10]Hoffmann, R., Zhang, C., Ling, X., et al., 2011. Knowledge-based weak supervision for information extraction of overlapping relations. Proc. 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies, p.541-550.

[11]Jenatton, R., Roux, N.L., Bordes, A., et al., 2012. A latent factor model for highly multi-relational data. NIPS, p.3167-3175.

[12]Lin, Y.K., Liu, Z.Y., Luan, H.B., et al., 2015. Modeling relation paths for representation learning of knowledge bases. arXiv: 1506.00379. http://arxiv.org/abs/1506.00379

[13]Maaten, L.V.D., Hinton, G., 2008. Visualizing data using t-SNE. J. Mach. Learn. Res., 9:2579-2605.

[14]Miller, G.A., 1995. WordNet: a lexical database for English. Commun. ACM, 38(11):39-41.

[15]Nickel, M., Tresp, V., Kriegel, H.P., 2011. A three-way model for collective learning on multi-relational data. Proc. 28th Int. Conf. on Machine Learning, p.809-816.

[16]Singhal, A., 2012. Introducing the Knowledge Graph: Things, not Strings. Google. https://21stcenturylibrary.com/2012/05/21/introducing-googles-knowledge-graph-things-not-strings [Accessed on May 21, 2012].

[17]Socher, R., Chen, D., Manning, C.D., et al., 2013. Reasoning with neural tensor networks for knowledge base completion. NIPS, p.926-934.

[18]Wang, Z., Zhang, J.W., Feng, J.L., et al., 2014. Knowledge graph embedding by translating on hyperplanes. Proc. 28th Conf. on Artificial Intelligence, p.1112-1119.

[19]Waters, R., 2012. Google to unveil search results overhaul. Financial Times, May.