TY - GEN
T1 - Memoreader
T2 - 2018 Conference on Empirical Methods in Natural Language Processing, EMNLP 2018
AU - Back, Seohyun
AU - Yu, Seunghak
AU - Indurthi, Sathish
AU - Kim, Jihie
AU - Choo, Jaegul
N1 - Funding Information:
We thank to reviewers for helpful feedback. This work is the extended version of Yu et al. (2018b). This work is partially supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIP) (No. NRF2016R1C1B2015924).
Publisher Copyright:
© 2018 Association for Computational Linguistics
PY - 2020/1/1
Y1 - 2020/1/1
N2 - Machine reading comprehension helps machines learn to utilize most of the human knowledge written in the form of text. Existing approaches made a significant progress comparable to human-level performance, but they are still limited in understanding, up to a few paragraphs, failing to properly comprehend lengthy document. In this paper, we propose a novel deep neural network architecture to handle a long-range dependency in RC tasks. In detail, our method has two novel aspects: (1) an advanced memory-augmented architecture and (2) an expanded gated recurrent unit with dense connections that mitigate potential information distortion occurring in the memory. Our proposed architecture is widely applicable to other models. We have performed extensive experiments with well-known benchmark datasets such as TriviaQA, QUASAR-T, and SQuAD. The experimental results demonstrate that the proposed method outperforms existing methods, especially for lengthy documents.
AB - Machine reading comprehension helps machines learn to utilize most of the human knowledge written in the form of text. Existing approaches made a significant progress comparable to human-level performance, but they are still limited in understanding, up to a few paragraphs, failing to properly comprehend lengthy document. In this paper, we propose a novel deep neural network architecture to handle a long-range dependency in RC tasks. In detail, our method has two novel aspects: (1) an advanced memory-augmented architecture and (2) an expanded gated recurrent unit with dense connections that mitigate potential information distortion occurring in the memory. Our proposed architecture is widely applicable to other models. We have performed extensive experiments with well-known benchmark datasets such as TriviaQA, QUASAR-T, and SQuAD. The experimental results demonstrate that the proposed method outperforms existing methods, especially for lengthy documents.
UR - http://www.scopus.com/inward/record.url?scp=85078309621&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85078309621&partnerID=8YFLogxK
M3 - Conference contribution
AN - SCOPUS:85078309621
T3 - Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, EMNLP 2018
SP - 2131
EP - 2140
BT - Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, EMNLP 2018
A2 - Riloff, Ellen
A2 - Chiang, David
A2 - Hockenmaier, Julia
A2 - Tsujii, Jun'ichi
PB - Association for Computational Linguistics
Y2 - 31 October 2018 through 4 November 2018
ER -