Abstract
Reproducibility is one of the pillars of scientific research. This study attempts to reproduce the Gated Self-Matching Network, which is the basis of one of the best performing models on the SQuAD dataset. We reimplement the neural network model and highlight ambiguities in the original architectural description. We show that due to uncertainty about only two components of the neural network model and no precise description of the training process, it is not possible to reproduce the experimental results obtained by the original implementation. Finally we summarize what we learned from this reproduction process about writing precise neural network architecture descriptions, providing our implementation as a basis for future exploration.
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.
Code available at https://github.com/alexduer/squad-gated-rep.
References
Chen, D., Bolton, J., Manning, C.D.: A thorough examination of the CNN/daily mail reading comprehension task. In: Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, pp. 2358–2367 (2016)
Cho, K., Van Merriënboer, B., Gulcehre, C., Bahdanau, D., Bougares, F., Schwenk, H., Bengio, Y.: Learning phrase representations using RNN encoder-decoder for statistical machine translation. In: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 1724–1734 (2014)
Glorot, X., Bengio, Y.: Understanding the difficulty of training deep feedforward neural networks. In: Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, pp. 249–256 (2010)
Hermann, K.M., Kocisky, T., Grefenstette, E., Espeholt, L., Kay, W., Suleyman, M., Blunsom, P.: Teaching machines to read and comprehend. In: Advances in Neural Information Processing Systems, pp. 1693–1701 (2015)
Hochreiter, S., Schmidhuber, J.: Long short-term memory. Neural Comput. 9(8), 1735–1780 (1997)
Hu, M., Peng, Y., Qiu, X.: Mnemonic reader for machine comprehension. arXiv preprint arXiv:1705.02798 (2017)
Kingma, D., Ba, J.: Adam: a method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014)
Manning, C.D., Surdeanu, M., Bauer, J., Finkel, J.R., Bethard, S., McClosky, D.: The stanford CoreNLP natural language processing toolkit. In: ACL (System Demonstrations), pp. 55–60 (2014)
Pennington, J., Socher, R., Manning, C.D.: Glove: global vectors for word representation. In: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), vol. 14, pp. 1532–1543 (2014)
Rajpurkar, P., Zhang, J., Lopyrev, K., Liang, P.: Squad: 100,000+ questions for machine comprehension of text. In: Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing (EMNLP) (2016)
Richardson, M., Burges, C.J., Renshaw, E.: Mctest: a challenge dataset for the open-domain machine comprehension of text. In: Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing (EMNLP), vol. 3, p. 4 (2013)
Rojas, R.: Neural Networks: A Systematic Introduction. Springer, Heidelberg (2013), https://doi.org/10.1007/978-3-642-61068-4
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)
Vinyals, O., Fortunato, M., Jaitly, N.: Pointer networks. In: Cortes, C., Lawrence, N.D., Lee, D.D., Sugiyama, M., Garnett, R. (eds.) Advances in Neural Information Processing Systems, vol. 28, pp. 2692–2700. Curran Associates, Inc. (2015)
Wang, S., Jiang, J.: Learning natural language inference with LSTM. In: Proceedings of the 15th Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (2016)
Wang, W., Yang, N., Wei, F., Chang, B., Zhou, M.: Gated self-matching networks for reading comprehension and question answering. In: Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (2017)
Xiong, C., Zhong, V., Socher, R.: Dynamic coattention networks for question answering. arXiv preprint arXiv:1611.01604 (2016)
Zeiler, M.D.: Adadelta: an adaptive learning rate method. arXiv preprint arXiv:1212.5701 (2012)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG, part of Springer Nature
About this paper
Cite this paper
Dür, A., Rauber, A., Filzmoser, P. (2018). Reproducing a Neural Question Answering Architecture Applied to the SQuAD Benchmark Dataset: Challenges and Lessons Learned. In: Pasi, G., Piwowarski, B., Azzopardi, L., Hanbury, A. (eds) Advances in Information Retrieval. ECIR 2018. Lecture Notes in Computer Science(), vol 10772. Springer, Cham. https://doi.org/10.1007/978-3-319-76941-7_8
Download citation
DOI: https://doi.org/10.1007/978-3-319-76941-7_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-76940-0
Online ISBN: 978-3-319-76941-7
eBook Packages: Computer ScienceComputer Science (R0)