文献类型：期刊文献

英文题名：Adverse Drug Reaction Related Post Detecting Using Sentiment Feature

作者：Liu, Jingfang[1];Jiang, Xiaoyan[1];Chen, Qiangyuan[2];Song, Mei[3];Li, Jia[4]

机构：[1]Shanghai Univ, Sch Management, Dept Informat Management, Shanghai 200444, Peoples R China;[2]Shanghai Univ, Sch Econ, Dept Econ, Shanghai 200444, Peoples R China;[3]Jiangsu Normal Univ, Sch Smart Educ, Dept Software Engn, Xuzhou 221116, Jiangsu, Peoples R China;[4]East China Univ Sci & Technol, Sch Business, Dept Management Sci & Engn, Shanghai 200237, Peoples R China

年份：2018

卷号：47

期号：6

起止页码：861

外文期刊名：IRANIAN JOURNAL OF PUBLIC HEALTH

收录：;WOS:【SSCI(收录号:WOS:000435672200012)，SCI-EXPANDED(收录号:WOS:000435672200012)】；

语种：英文

外文关键词：Adverse drug reaction; Post; Sentiment feature

摘要：Background: The posts related to Adverse Drug Reaction (ADR) on social websites are believed to be valuable resource for post-marketing drug surveillance. Beyond domain feature, the aim of this study was to find a more effective method to detect ADR related post. Methods: We conducted experiment on posts using sentiment features from March 8 to May 20 in 2016 in Shanghai of China. Firstly, the diabetes posts were collected; the 1814 posts were annotated by hand. Secondly, sentiment features set were generated and the chi(2). (CHI) statistics were used to select feature. Finally, we evaluated the effectiveness of our method using the different feature sets. Results: By comparing the posts detection performance of different feature sets, using sentiment features by CHI statistics can improve ADR related post detection performance. By comparing the ADR-related group with the non-ADR group, performance of ADR related post detection was better than the performance of nonADR post detection. We could obtain highest performance owing to introducing sentiment feature and using CHI feature selection technique, and the method was proved to be effective during detecting post related to ADR. Conclusion: By using sentiment feature and CHI feature selection technique, we can get an effective method to detect post related to ADR.