Investors obtain data or share information through the Indonesia Stock Exchange web page. From online information portals, investors occasionally obtain more knowledge about stock analysis and profitable stock projections. However, it takes time for investors to identify subjects that frequently come up and become popular stock information themes. The core of online stock information therefore requires subject modeling. Topic modeling, which seeks to look at the issues being discussed on a page of an internet stock information website, is the focus of this study. Research information is combined with 181 stock information February to July 2021, web source: kontan.co.id. Utilizing the Latent Dirichlet Allocation approach, topic modeling was tested. The topic modeling created 3 important groups and the distribution of the 10 creator words of each topic from online stock information. The topic modeling description creates a thirty-word note that occurs frequently from each topic.
Copyrights © 2023