Answer Quality Evaluation in Online Health Care Community

DOI

10.2991/ncce-18.2018.143How to use a DOI?

Keywords

evaluation rules; QA; promote the f1-score.

Abstract

Nowadays, through the online health care community platform, users can raise health related questions and doctors would provide corresponding answers. However, some answers could be low-quality and repeated. In this paper, we aim to evaluate and predict the quality of the answers in online health care communities. We set the evaluation rules and scoring model for the medical answer text. 12 features are proposed to represent the answer quality. 8 classic classification models are used to predict the answer score. The best model get 0.90 f1-score. Furthermore, we utilize our model to select QA data of high quality, which help the QA matching task and promote the f1-score from 0.86 to 0.93.

Copyright

© 2018, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Binjun Zhu
AU  - Xiaofeng Cai
AU  - Ruichu Cai
PY  - 2018/05
DA  - 2018/05
TI  - Answer Quality Evaluation in Online Health Care Community
BT  - Proceedings of the 2018 International Conference on Network, Communication, Computer Engineering (NCCE 2018)
PB  - Atlantis Press
SP  - 864
EP  - 868
SN  - 1951-6851
UR  - https://doi.org/10.2991/ncce-18.2018.143
DO  - 10.2991/ncce-18.2018.143
ID  - Zhu2018/05
ER  -