Elamur Research
☰
home / papers / 2020 / Improving Multi-hop Question Answering over Knowledge Graphs using Knowledge Base Embeddings

Improving Multi-hop Question Answering over Knowledge Graphs using Knowledge Base Embeddings

Written for the Research library (claude-opus-5-5) from the paper text. Original — aclanthology.org/2020.acl-main.412.pdf

Summary

The paper addresses multi-hop KGQA: answering a natural-language question over a knowledge graph when several edges must be traversed. Real-world KGs are often incomplete. Because of missing links, the path to the answer becomes longer or disappears altogether. Earlier methods compensated for this with external text corpora, which are not always available. In addition, they searched for the answer only within a pre-restricted neighborhood of the node. The authors propose EmbedKGQA, the first method to apply KG embeddings (ComplEx) to this task. The system has three modules. The first learns vector representations of the graph's entities and relations, the second maps the question to a vector in the same space, and the third scores all entities as candidate answers. Relation similarity helps select the best one. Since the embeddings can predict missing links, the model finds the answer even when the graph has no direct path to it, and without additional data. Experiments on MetaQA and WebQuestionsSP show an advantage over strong baselines, especially on sparse graphs. This matters for incomplete knowledge bases, including those for lower-resource languages.

Key points

  • Method: ComplEx for the graph, a question vector in complex space, and scoring of the triple (head entity, question, candidate). No annotation of question type or reasoning steps is required.
  • Test on an incomplete graph: from MetaQA 1-hop, the triples that directly give the answer to the validation questions, as well as their paraphrases, were removed. Subgraph-based methods theoretically get 0 hits@1 here, while EmbedKGQA gets 29.9. Plain tail prediction with ComplEx embeddings gives only 20.1.
  • The WebQSP graph (a subset of Freebase) contains about 1.8M entities, MetaQA about 134k.
  • In ablations on WebQSP, disabling the relation matching module noticeably worsens the result. Restricting candidates to the 2-hop neighborhood helps on the full graph but hurts on the sparse KG-50.
  • The code is open: github.com/malllabiisc/EmbedKGQA.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: AI & machine learning

Cited by

165 papers of our library cite it (of the 140673 whose bibliographies we read); 3 of them have a page here, 162 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2020.acl-main.412

@inproceedings{saxena2020improving,
  title = {{Improving Multi-hop Question Answering over Knowledge Graphs using Knowledge Base Embeddings}},
  author = {Saxena, Apoorv and Tripathi, Aditay and Talukdar, Partha},
  booktitle = {Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics},
  year = {2020},
  doi = {10.18653/v1/2020.acl-main.412},
  url = {https://doi.org/10.18653/v1/2020.acl-main.412}
}

References

36 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Ivana Balažević, Carl Allen, and Timothy M Hospedales. 2019. Tucker: Tensor factorization for knowledge graph completion. arXiv preprint arXiv:1901.09590. find our card
  2. Junwei Bao, Nan Duan, Zhao Yan, Ming Zhou, and Tiejun Zhao. 2016. Constraint-based question answering with knowledge graph. In Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, pages 2503–2514. find
  3. Antoine Bordes, Sumit Chopra, and Jason Weston. Sören Auer, et al. 2015. Dbpedia–a large-scale, multilingual knowledge base extracted from wikipedia. Semantic Web, 6(2):167–195. 2014a. Question answering with subgraph embeddings. arXiv preprint arXiv:1406.3676. find
  4. Antoine Bordes, Nicolas Usunier, Sumit Chopra, and Jason Weston. 2015. Large-scale simple question answering with memory networks. arXiv preprint arXiv:1506.02075. find
  5. Antoine Bordes, Nicolas Usunier, Alberto Garcia-Duran, Jason Weston, and Oksana Yakhnenko. 2013. Translating embeddings for modeling multi-relational data. In Advances in neural information processing systems, pages 2787–2795. find
  6. Antoine Bordes, Jason Weston, and Nicolas Usunier. 2014b. Open question answering with weakly supervised embedding models. In Joint European conference on machine learning and knowledge discovery in databases, pages 165–180. Springer. find
  7. Zihang Dai, Lei Li, and Wei Xu. 2016. Cfo: Conditional focused neural question answering with large-scale knowledge bases. arXiv preprint arXiv:1606.01994. find
  8. Tim Dettmers, Pasquale Minervini, Pontus Stenetorp, and Sebastian Riedel. 2018. Convolutional 2d knowledge graph embeddings. In Thirty-Second AAAI Conference on Artificial Intelligence. find
  9. Li Dong, Furu Wei, Ming Zhou, and Ke Xu. 2015. Question answering over freebase with multi-column convolutional neural networks. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volum… find
  10. Freebase data dumps. https:// developers.google.com/freebase/data . find
  11. Yanchao Hao, Yuanzhe Zhang, Kang Liu, Shizhu He, Zhanyi Liu, Hua Wu, and Jun Zhao. 2017. An end-to-end model for question answering over knowledge base with cross-attention combining global knowledge. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume… find
  12. Frank L Hitchcock. 1927. The expression of a tensor or a polyadic as a sum of products. Journal of Mathematics and Physics, 6(1-4):164–189. find
  13. Seyed Mehran Kazemi and David Poole. 2018. Simple embedding for link prediction in knowledge graphs. In Advances in Neural Information Processing Systems, pages 4284–4295. find
  14. Thomas N Kipf and Max Welling. 2016. Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907. find our card
  15. Jens Lehmann, Robert Isele, Max Jakob, Anja Jentzsch, Dimitris Kontokostas, Pablo N Mendes, Sebastian Hellmann, Mohamed Morsey, Patrick Van Kleef, Dingcheng Li, Jingyuan Zhang, and Ping Li. 2018. Representation learning for question classification via topic sparse autoencoder and entity embedding. I… find
  16. Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692. find
  17. Denis Lukovnikov, Asja Fischer, Jens Lehmann, and Sören Auer. 2017. Neural network-based question answering over knowledge graphs on word and character level. In Proceedings of the 26th international conference on World Wide Web, pages 1211–1220. International World Wide Web Conferences Steering Co… find
  18. Alexander Miller, Adam Fisch, Jesse Dodge, Amir-Hossein Karimi, Antoine Bordes, and Jason Weston. 2016. Key-value memory networks for directly reading documents. arXiv preprint arXiv:1606.03126. find
  19. Tom M. Mitchell, William W. Cohen, Estevam R. Hruschka Jr., Partha P. Talukdar, Bo Yang, Justin Betteridge, Andrew Carlson, Bhavana Dalvi Mishra, Matt Gardner, Bryan Kisiel, Jayant Krishnamurthy, Ni Lao, Kathryn Mazaitis, Thahir Mohamed, Ndapandula Nakashole, Emmanouil A. Platanios, Alan Ritter, Meh… find
  20. Salman Mohammed, Peng Shi, and Jimmy Lin. 2017. Strong baselines for simple question answering over knowledge graphs with and without neural networks. arXiv preprint arXiv:1712.01969. find
  21. Maximilian Nickel, Volker Tresp, and Hans-Peter Kriegel. 2011. A three-way model for collective learning on multi-relational data. In ICML, volume 11, pages 809–816. find
  22. Fabian M Suchanek, Gjergji Kasneci, and Gerhard Weikum. 2007. Yago: a core of semantic knowledge. In Proceedings of the 16th international conference on World Wide Web, pages 697–706. ACM. find
  23. Haitian Sun, Tania Bedrax-Weiss, and William W Cohen. 2019a. Pullnet: Open domain question answering with iterative retrieval on knowledge bases and text. arXiv preprint arXiv:1904.09537. find
  24. Haitian Sun, Bhuwan Dhingra, Manzil Zaheer, Kathryn Mazaitis, Ruslan Salakhutdinov, and William W Cohen. 2018. Open domain question answering using early fusion of knowledge bases and text. arXiv preprint arXiv:1809.00782. find
  25. Zhiqing Sun, Zhi-Hong Deng, Jian-Yun Nie, and Jian Tang. 2019b. Rotate: Knowledge graph embedding by relational rotation in complex space. arXiv preprint arXiv:1902.10197. find
  26. Théo Trouillon, Johannes Welbl, Sebastian Riedel, Éric Gaussier, and Guillaume Bouchard. 2016. Complex embeddings for simple link prediction. In International Conference on Machine Learning, pages 2071–2080. find
  27. Ledyard R Tucker. 1966. Some mathematical notes on three-mode factor analysis. Psychometrika, 31(3):279–311. find
  28. Ferhan Ture and Oliver Jojic. 2016. No need to pay attention: Simple recurrent neural networks work!(for answering” simple” questions). arXiv preprint arXiv:1606.05029. find
  29. Shikhar Vashishth, Soumya Sanyal, Vikram Nitin, Nilesh Agrawal, and Partha Talukdar. 2019. Interacte: Improving convolution-based knowledge graph embeddings by increasing feature interactions. arXiv preprint arXiv:1911.00219. — [arXiv:1911.00219](https://arxiv.org/abs/1911.00219) find
  30. Bishan Yang, Wen-tau Yih, Xiaodong He, Jianfeng Gao, and Li Deng. 2014a. Embedding entities and relations for learning and inference in knowledge bases. arXiv preprint arXiv:1412.6575. find
  31. Min-Chul Yang, Nan Duan, Ming Zhou, and Hae-Chang Rim. 2014b. Joint relational embeddings for knowledge-based question answering. In Proceedings of the 2014 conference on empirical methods in natural language processing (EMNLP), pages 645– 650. find
  32. Min-Chul Yang, Do-Gil Lee, So-Young Park, and Hae-Chang Rim. 2015. Knowledge-based question answering using the semantic embedding space. Expert Systems with Applications, 42(23):9086–9104. find
  33. Scott Wen-tau Yih, Ming-Wei Chang, Xiaodong He, and Jianfeng Gao. 2015. Semantic parsing via staged query graph generation: Question answering with knowledge base. find
  34. Wen tau Yih, Matthew Richardson, Christopher Meek, Ming-Wei Chang, and Jina Suh. 2016. The value of semantic parse labeling for knowledge base question answering. In ACL. find
  35. Wenpeng Yin, Mo Yu, Bing Xiang, Bowen Zhou, and Hinrich Schütze. 2016. Simple question answering by attentive convolutional neural network. arXiv preprint arXiv:1606.03391. find
  36. Yuyu Zhang, Hanjun Dai, Zornitsa Kozareva, Alexander J Smola, and Le Song. 2018. Variational reasoning for question answering with knowledge graph. In Thirty-Second AAAI Conference on Artificial Intelligence. find