Abstract
Multi-hop machine reading comprehension (MRC) requires models to mine and utilize relevant information from multiple documents to predict the answer to a semantically related question. Existing work resorts to either document-level or entity-level inference among relevant information, which can be too coarse or too subtle, resulting less accurate understanding of the texts. To mitigate the issue, this research proposes a sentence-based multi-hop reasoning approach named SMR. SMR starts with sentences of documents, and unites the question to establish several reasoning chains based on sentence-level representations. In addition, to resolve the complication of pronouns on sentence semantics, we concatenate two sentences, if necessary, to assist in constructing reasoning chains. The model then synthesizes the information existed in all the reasoning chains, and predicts a probability distribution for selecting the correct answer. In experiments, we evaluate SMR on two popular multi-hop MRC benchmark datasets - WikiHop and MedHop. The model achieves 68.3 and 62.9 in terms of accuracy, respectively, exhibiting a remarkable improvement over state-of-the-art option. Additionally, qualitative analysis also demonstrates the validity and interpretability of SMR.
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
Notes
- 1.
We are in the process of obtaining the results on the hidden test set.
References
Nishida, K., et al.: Multi-style generative reading comprehension. ACL (1), 2273–2284 (2019)
Wang, C., Jiang, H.: Explicit utilization of general knowledge in machine reading comprehension. ACL (1), 2263–2272 (2019)
Jiang, Y., Joshi, N., Chen, Y.-C., Bansal, M.: Explore, propose, and assemble: an interpretable model for multi-hop reading comprehension. ACL (1), 2714–2725 (2019)
Kundu, S., Khot, T., Sabharwal, A., Clark, P.: Exploiting explicit paths for multi-hop reading comprehension. ACL (1), 2737–2747 (2019)
Tu, M., Wang, G., Huang, J., Tang, Y., He, X., Zhou, B.: Multi-hop reading comprehension across multiple documents by reasoning over heterogeneous graphs. ACL (1), 2704–2713 (2019)
Ding, M., Zhou, C., Chen, Q., Yang, H., Tang, J.: Cognitive graph for multi-hop reading comprehension at scale. ACL (1), 2694–2703 (2019)
Zhong, V., Xiong, C., Keskar, N.S., Socher, R.: Coarse-grain fine-grain coattention network for multi-evidence question answering. ICLR (Poster) (2019)
Welbl, J., Stenetorp, P., Riedel, S.: Constructing datasets for multi-hop reading comprehension across documents. TACL 6, 287–302 (2018)
Yang, Z., et al.: HotpotQA: a dataset for diverse, explainable multi-hop question answering. EMNLP, pp. 2369–2380 (2018)
Rajpurkar, P., Zhang, J., Lopyrev, K., Liang, P.: SQuAD: 100, 000+ questions for machine comprehension of text. EMNLP, pp. 2383–2392 (2016)
Devlin, J., Chang, M.-W., Lee, K., Toutanova, K.: BERT: Pre-training of deep bidirectional transformers for language understanding. NAACL-HLT (1), 4171–4186 (2019)
Pennington, J., Socher, R., Manning, C.D.: Glove: global vectors for word representation. EMNLP, 1532–1543 (2014)
De Cao, N., Aziz, W., Titov, I.: Question answering by reasoning across documents with graph convolutional networks. NAACL-HLT (1), 2306–2317 (2019)
Cho, K., et al.: Learning phrase representations using RNN encoder-decoder for statistical machine translation. EMNLP, pp. 1724–1734 (2014)
Bird, S., Loper, E.: NLTK: the natural language toolkit. ACL (Poster and Demonstration) (2004)
Peters, M.E., et al.: Deep contextualized word representations. NAACL-HLT, pp. 2227–2237 (2018)
Joshi, M., Choi, E., Weld, D.S., Zettlemoyer, L.: TriviaQA: a large scale distantly supervised challenge dataset for reading comprehension. ACL (1), pp. 1601–1611 (2017)
Onishi, T., Wang, H., Bansal, M., Gimpel, K., McAllester, D.A.: Who did what: a large-scale person-centered cloze dataset. EMNLP, pp. 2230–2235 (2016)
Hill, F., Bordes, A., Chopra, S., Weston, J.: The goldilocks principle: reading children’s books with explicit memory representations. ICLR (2016)
Battaglia, P.W., et al.: Relational inductive biases, deep learning, and graph networks. CoRR abs/1806.01261 (2018)
Srivastava, R.K., Greff, K., Schmidhuber, J.: Highway Networks. CoRR abs/1505.00387 (2015)
Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural Comput. 9(8), 1735–1780 (1997)
Chandar, S., Ahn, S., Larochelle, H., Vincent, P., Tesauro, G., Bengio, Y.: Hierarchical Memory Networks. CoRR abs/1605.07427 (2016)
Bahdanau, D., Cho, K., Bengio, Y.: Neural machine translation by jointly learning to align and translate. ICLR (2015)
Srivastava, N., Hinton, G.E., Krizhevsky, A., Sutskever, I., Salakhutdinov, R.: Dropout: a simple way to prevent neural networks from overfitting. J. Mach. Learn. Res. 15(1), 1929–1958 (2014)
Seo, M., Kembhavi, A., Farhadi, A., Hajishirzi, H.: Bidirectional attention flow for machine comprehension. ICLR (2017)
Cao, Y., Fang, M., Tao, D.: BAG: bi-directional attention entity graph convolutional network for multi-hop reasoning question answering. NAACL-HLT (1), 357–362 (2019)
Acknowledgement
This work was partially supported by NSFC under grants Nos. 61872446, 61902417 and 71971212, and PNSF of Hunan under grant No. 2019JJ20024.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Switzerland AG
About this paper
Cite this paper
Huo, L., Ge, B., Zhao, X. (2020). Multi-hop Reading Comprehension Incorporating Sentence-Based Reasoning. In: Wang, X., Zhang, R., Lee, YK., Sun, L., Moon, YS. (eds) Web and Big Data. APWeb-WAIM 2020. Lecture Notes in Computer Science(), vol 12317. Springer, Cham. https://doi.org/10.1007/978-3-030-60259-8_40
Download citation
DOI: https://doi.org/10.1007/978-3-030-60259-8_40
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-60258-1
Online ISBN: 978-3-030-60259-8
eBook Packages: Computer ScienceComputer Science (R0)