Abstract
Multi-document question answering is a hot-spot research task that retrieves answers for a natural language question from a set of documents. Although existing models try to enhance the recall ability, the performance of answering short questions is still far from satisfaction. In this paper, we address multi-document question answering by leveraging external knowledge to assist the semantic understanding, improving the overall performance. Specifically, we learn the relationship between questions and candidate paragraphs, and select valuable external knowledge. The machine reading comprehension model is trained to predict the answers in the screened external knowledge, which verifies the role of external knowledge in supplementing the answers. Then keywords are extracted to assist the indexing ability of answering short queries over multiple documents. Gate and Attention components are designed to integrate external knowledge through the deep neural network, assisting the retrieval of answers. We investigate the role of external knowledge in supplementing answers, assisting recall, and automatically assisting of deep neural network in the multi-document question answering system. Experimental results confirm the effectiveness of the proposed method.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Asakiewicz, C., Stohr, E.A., Mahajan, S., Pandey, L.: Building a cognitive application using watson deepqa. IT Prof. 19(4), 36–44 (2017)
Broens, T.H.F., Pokraev, S., van Sinderen, M., Koolwaaij, J., Costa, P.D.: Context-aware, ontology-based service discovery. In: EUSAI, vol. 3295, pp. 72–83 (2004)
Devlin, J., Chang, M., Lee, K., Toutanova, K.: BERT: pre-training of deep bidirectional transformers for language understanding. In: NAACL-HLT, pp. 4171–4186 (2019)
Glass, M.R., et al.: Span selection pre-training for question answering. In: ACL, pp. 2773–2782 (2020)
Goddeau, D., Meng, H.M., Polifroni, J., Seneff, S., Busayapongchai, S.: A form-based dialogue manager for spoken language applications. In: The 4th International Conference on Spoken Language Processing (1996)
Group, N.: R-net: machine reading comprehension with self-matching networks (2017)
Hao, X., Chang, X., Liu, K.: A rule-based Chinese question answering system for reading comprehension tests. In: IIH-MSP, pp. 325–329 (2007)
Huang, H., Zhu, C., Shen, Y., Chen, W.: Fusionnet: fusing via fully-aware attention with application to machine comprehension. In: ICLR (2018)
Kruth, J., Ginderachter, T.V., Tanaya, P.I., Valckenaers, P.: The use of finite state machines for task-based machine tool control. Comput. Ind. 46(3), 247–258 (2001)
Lee, C., Cha, Y.S., Kuc, T.Y.: Implementation of dialogue system for intelligent service robots. In: 2008 International Conference on Control, Automation and Systems (2008)
Lu, J., Ábrego, G.H., Ma, J., Ni, J., Yang, Y.: Multi-stage training with improved negative contrast for neural passage retrieval. In: EMNLP, pp. 6091–6103 (2021)
Qu, Y., et al.: Rocketqa: an optimized training approach to dense passage retrieval for open-domain question answering. In: NAACL-HLT, pp. 5835–5847
Sales, J.E., Freitas, A., Handschuh, S.: An open vocabulary semantic parser for end-user programming using natural language. In: ICSC, pp. 77–84 (2018)
Seo, M.J., Kembhavi, A., Farhadi, A., Hajishirzi, H.: Bidirectional attention flow for machine comprehension. In: ICLR (2017)
Shen, Y., He, X., Gao, J., Deng, L., Mesnil, G.: Learning semantic representations using convolutional neural networks for web search. In: WWW, pp. 373–374 (2014)
Sun, C., Qiu, X., Xu, Y., Huang, X.: How to fine-tune BERT for text classification? In: CCL, pp. 194–206 (2019)
Wang, Y., et al.: Multi-passage machine reading comprehension with cross-passage answer verification. In: ACL, pp. 1918–1927 (2018)
Wei, Z., Xuan, Z., Chen, J.: Design and implementation of influenza question answering system based on multi-strategies. In: CSAE (2012)
Wu, Y., Nong, G., Chan, W.H., Han, L.B.: Checking big suffix and LCP arrays by probabilistic methods. IEEE Trans. Comput. 66(10), 1667–1675 (2017)
Xiong, C., Zhong, V., Socher, R.: Dynamic coattention networks for question answering. In: ICLR (2017)
Yu, J., et al.: Modelling domain relationships for transfer learning on retrieval-based question answering systems in e-commerce. In: WSDM, pp. 682–690 (2018)
Zhang, J., Zhu, X., Chen, Q., Dai, L., Wei, S., Jiang, H.: Exploring question understanding and adaptation in neural-network-based question answering. CoRR (2017)
Zheng, B., et al.: Document modeling with graph attention networks for multi-grained machine reading comprehension. In: ACL, pp. 6708–6718 (2020)
Acknowledgement
This work was supported by National Natural Science Foundation of China (Grant No. 61902074) and Science and Technology Committee Shanghai Municipality (Grant No. 19ZR1404900).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Yan, Z., Zheng, W. (2022). Multi-document Question Answering Powered by External Knowledge. In: Chbeir, R., Huang, H., Silvestri, F., Manolopoulos, Y., Zhang, Y. (eds) Web Information Systems Engineering – WISE 2022. WISE 2022. Lecture Notes in Computer Science, vol 13724. Springer, Cham. https://doi.org/10.1007/978-3-031-20891-1_33
Download citation
DOI: https://doi.org/10.1007/978-3-031-20891-1_33
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-20890-4
Online ISBN: 978-3-031-20891-1
eBook Packages: Computer ScienceComputer Science (R0)