A Study of Online Health Community Medication Counseling Based on Text Mining postprint

Abstract: ［Purpose/Significance］Online health community has become an important way for people to obtain health information, and studying the demand of medication consultation of online health community users contributes to the optimization and sustainable development of online health community drug service. ［Method/Process］Taking 39Health.com as an example, firstly, 59,048 medication consultation comments of gastrointestinal medicines are crawled using Python coding and pre-processed; secondly, the experimental data are subjected to theme keyword mining using text mining methods such as TF-IDF, TextRank, and LDA topic model, and keyword co-occurrence network analysis is carried out; lastly, a comprehensive analysis of the online health community users’ medication consulting needs of the theme characteristics, and put forward optimization suggestions. ［Result/Conclusion］ The results of this study show that online health community users are mainly concerned about the therapeutic effect of drugs, the way of taking drugs, adverse reactions, the difference between drugs, and the precautions to be taken during pregnant women and other special groups using drugs. This study provides a theoretical basis for drug manufacturers to adjust and optimize the content of drug manuals, and on the other hand, it provides a direction for online health communities to optimize the content layout of drug manuals and to establish or improve drug popularization services.