@inproceedings{2fdad90309af402099be2120c8d5366e,
title = "Tea in benefits of health: A literature analysis using text mining and latent dirichlet allocation",
abstract = "Tea originated in Asian, which was initially used as a medicinal herb. The variety of tea is according to different manufacturing processes and levels of oxidation. The different varieties of tea have different level of effects on health, thus this study adopted text mining technique and Latent Dirichlet Allocation (LDA) to analyze literature for tea in health effect. This study chose Web of Science as the database of literature source, and the search literature from 2007 to 2017. The total 1230 journal articles were collected in this study. The title, abstract, and keywords of the collected journal articles were used as a dataset for the experiment. Experimental results show that the VEM method is significantly lower than Gibbs sampling in perplexity. Hence, this study chooses K=150 when VEM method and Gibbs sampling reach the minimal perplexity in the same time. Many topics that related with tea and compounds of tea, however some topics had terms that related to health and disease. The top 10 topics show that tea could reduce the risk of diseases and benefit of health.",
keywords = "Health, Latent Dirichlet Allocation, LDA, Literature analysis, Tea, Text mining, Topic model",
author = "Cheng, {Ching Hsue} and Hung, {Wei Lun}",
note = "Publisher Copyright: {\textcopyright} 2018 Association for Computing Machinery.; 2nd International Conference on Medical and Health Informatics, ICMHI 2018 ; Conference date: 08-06-2018 Through 10-06-2018",
year = "2018",
month = jun,
day = "8",
doi = "10.1145/3239438.3239459",
language = "English",
series = "ACM International Conference Proceeding Series",
publisher = "Association for Computing Machinery (ACM)",
pages = "148--155",
booktitle = "ICMHI 2018 - Proceedings of 2018 the 2nd International Conference on Medical and Health Informatics",
address = "United States",
}
