Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Jun 5;20(6):439.
doi: 10.3390/e20060439.

Factoid Question Answering with Distant Supervision

Hongzhi Zhang  1   2 , Xiao Liang  1 , Guangluan Xu  1   2 , Kun Fu  1   2   3 , Feng Li  1 , Tinglei Huang  1
Affiliations

Factoid Question Answering with Distant Supervision

Hongzhi Zhang et al. Entropy (Basel). .

Abstract

Automatic question answering (QA), which can greatly facilitate the access to information, is an important task in artificial intelligence. Recent years have witnessed the development of QA methods based on deep learning. However, a great amount of data is needed to train deep neural networks, and it is laborious to annotate training data for factoid QA of new domains or languages. In this paper, a distantly supervised method is proposed to automatically generate QA pairs. Additional efforts are paid to let the generated questions reflect the query interests and expression styles of users by exploring the community QA. Specifically, the generated questions are selected according to the estimated probabilities they are asked. Diverse paraphrases of questions are mined from community QA data, considering that the model trained on monotonous synthetic questions is very sensitive to variants of question expressions. Experimental results show that the model solely trained on generated data via the distant supervision and mined paraphrases could answer real-world questions with the accuracy of 49.34%. When limited annotated training data is available, significant improvements could be achieved by incorporating the generated data. An improvement of 1.35 absolute points is still observed on WebQA, a dataset with large-scale annotated training samples.

Keywords: distant supervision; question answering; question paraphrase; reading comprehension.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
The structure of the QA model.
Figure 2
Figure 2
Statistics of WebQA and training data generated via distant supervision.
Figure 3
Figure 3
Distribution of the paraphrased predicates.
Figure 4
Figure 4
Word length distribution.
Figure 5
Figure 5
Factoid QA via distant supervision.
Figure 6
Figure 6
Improved factoid QA with distant supervision.
Figure 7
Figure 7
Curves of training loss and validation accuracy. SL denotes supervised learning. Pre-training+ SL denotes that the model is pre-trained on generated data and then trained on the annotated data. SL+ denotes the model simultaneously trained on generated data and annotated data.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Berant J., Chou A., Frostig R., Liang P. Semantic Parsing on Freebase from Question-Answer Pairs; Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing; Seattle, WA, USA. 18–21 October 2013; pp. 1533–1544.
    1. Bordes A., Usunier N., Chopra S., Weston J. Large-scale Simple Question Answering with Memory Networks. arXiv. 2015. 1506.02075
    1. Sun H., Ma H., He X., Yih W.t., Su Y., Yan X. Table Cell Search for Question Answering; Proceedings of the 25th International Conference on World Wide Web; Republic and Canton of Geneva, Switzerland. 11–15 April 2016; pp. 771–782.
    1. Rajpurkar P., Zhang J., Lopyrev K., Liang P. SQuAD: 100,000+ Questions for Machine Comprehension of Text. arXiv. 2016. 1606.05250
    1. Li P., Li W., He Z., Wang X., Cao Y., Zhou J., Xu W. Dataset and Neural Recurrent Sequence Labeling Model for Open-Domain Factoid Question Answering. arXiv. 2016. 1607.06275

LinkOut - more resources