Attentive Path Combination for Knowledge Graph Completion

Xiaotian Jiang, Quan Wang, Baoyuan Qi, Yongqin Qiu, Peng Li, Bin Wang
Proceedings of the Ninth Asian Conference on Machine Learning, PMLR 77:590-605, 2017.

Abstract

Knowledge graphs (KGs) are often significantly incomplete, necessitating a demand for KG completion. Path-based relation inference is one of the most important approaches to this task. Traditional methods treat each path between entity pairs as an atomic feature, thus inducing sparsity. Recently, neural network models solve this problem by decomposing a path as the sequence of relations in the path, before modelling path representations with Recurrent Neural Network (RNN) architectures. In cases there are multiple paths between an entity pair, state-of-the-art neural models either select only one path, or make usage of simple score pooling methods like Top-K, Average, LogSumExp. Unfortunately, none of these methods can model the scenario where relations can only be inferred by considering multiple informative paths collectively. In this paper, we propose a novel path-based relation inference model that learns entity pair representations with attentive path combination. Given an entity pair and a set of paths connecting the pair, our model allows for integrating information from each informative path, and form a dynamic entity pair representation for each query relation. We empirically evaluate the proposed method on a real-world dataset. Experimental results show that the proposed model achieves better performance than state-of-the-art path-based relation inference methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v77-jiang17a, title = {Attentive Path Combination for Knowledge Graph Completion}, author = {Jiang, Xiaotian and Wang, Quan and Qi, Baoyuan and Qiu, Yongqin and Li, Peng and Wang, Bin}, booktitle = {Proceedings of the Ninth Asian Conference on Machine Learning}, pages = {590--605}, year = {2017}, editor = {Zhang, Min-Ling and Noh, Yung-Kyun}, volume = {77}, series = {Proceedings of Machine Learning Research}, address = {Yonsei University, Seoul, Republic of Korea}, month = {15--17 Nov}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v77/jiang17a/jiang17a.pdf}, url = {https://proceedings.mlr.press/v77/jiang17a.html}, abstract = {Knowledge graphs (KGs) are often significantly incomplete, necessitating a demand for KG completion. Path-based relation inference is one of the most important approaches to this task. Traditional methods treat each path between entity pairs as an atomic feature, thus inducing sparsity. Recently, neural network models solve this problem by decomposing a path as the sequence of relations in the path, before modelling path representations with Recurrent Neural Network (RNN) architectures. In cases there are multiple paths between an entity pair, state-of-the-art neural models either select only one path, or make usage of simple score pooling methods like Top-K, Average, LogSumExp. Unfortunately, none of these methods can model the scenario where relations can only be inferred by considering multiple informative paths collectively. In this paper, we propose a novel path-based relation inference model that learns entity pair representations with attentive path combination. Given an entity pair and a set of paths connecting the pair, our model allows for integrating information from each informative path, and form a dynamic entity pair representation for each query relation. We empirically evaluate the proposed method on a real-world dataset. Experimental results show that the proposed model achieves better performance than state-of-the-art path-based relation inference methods.} }
Endnote
%0 Conference Paper %T Attentive Path Combination for Knowledge Graph Completion %A Xiaotian Jiang %A Quan Wang %A Baoyuan Qi %A Yongqin Qiu %A Peng Li %A Bin Wang %B Proceedings of the Ninth Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2017 %E Min-Ling Zhang %E Yung-Kyun Noh %F pmlr-v77-jiang17a %I PMLR %P 590--605 %U https://proceedings.mlr.press/v77/jiang17a.html %V 77 %X Knowledge graphs (KGs) are often significantly incomplete, necessitating a demand for KG completion. Path-based relation inference is one of the most important approaches to this task. Traditional methods treat each path between entity pairs as an atomic feature, thus inducing sparsity. Recently, neural network models solve this problem by decomposing a path as the sequence of relations in the path, before modelling path representations with Recurrent Neural Network (RNN) architectures. In cases there are multiple paths between an entity pair, state-of-the-art neural models either select only one path, or make usage of simple score pooling methods like Top-K, Average, LogSumExp. Unfortunately, none of these methods can model the scenario where relations can only be inferred by considering multiple informative paths collectively. In this paper, we propose a novel path-based relation inference model that learns entity pair representations with attentive path combination. Given an entity pair and a set of paths connecting the pair, our model allows for integrating information from each informative path, and form a dynamic entity pair representation for each query relation. We empirically evaluate the proposed method on a real-world dataset. Experimental results show that the proposed model achieves better performance than state-of-the-art path-based relation inference methods.
APA
Jiang, X., Wang, Q., Qi, B., Qiu, Y., Li, P. & Wang, B.. (2017). Attentive Path Combination for Knowledge Graph Completion. Proceedings of the Ninth Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 77:590-605 Available from https://proceedings.mlr.press/v77/jiang17a.html.

Related Material