With the rapid adoption of smart devices inrecent times, the mobile internet-based app market has witnessed significantgrowth across various sectors such as securities, education, and manufacturing.Generally, providers of mobile app services strive to address users' new demandsand challenges by updating their software. Analyzing user responses to updatesis crucial for successful maintenance. In this study, we propose a method toanalyze user reactions using topic modeling techniques applied to user reviews.To achieve this, from the Mobile Trading Systems (MTS) available on the Google Play Store, we selected “Kiwoom Securities Hero Moon S” due to itshighest number of reviews to gather review data. We employed the Latent Dirichlet Allocation (LDA) topic modeling technique to extract topics from thereviews and analyzed the trend of topic changes before and after app updates.The analysis revealed that the topic of connection errors, which dominated post-update reviews, showed a gradually stabilizing trend. However, the user authentication topic indicated an increase in diverse complaints after the app update, highlighting the need to actively address these issues. Additionally,in the case of MTS, the screen layout is very crucial to users, and numerous complaints arose when users' configured screens were reset. Therefore, it is evident that maintaining a consistent user interface is essential for usersatisfaction.
Read full abstract