The Business Value of Text Analysis and Topic Modeling: Evaluating Sentiment Shifts in Mobile Game Reviews after Updates

This article explores the potential of web scraping and text analysis as tools to extract business value from user reviews on Google Play. By collecting and analysing app reviews, the study investigates customer sentiment and perceptions before and after the release of specific app features. The proposed methodology leverages natural language processing (NLP) techniques to identify key trends and insights, providing actionable feedback for developers and stakeholders. This approach demonstrates how businesses can use real-time user feedback to assess feature performance, improve user satisfaction, and inform strategic decisions. A particular method used is splitting the data set in two subsets based on a specific date, trying to unveil potential shifts in users’ perceptions of the app before and after a major update. Bringing new perspectives to business value, a topic modeling analysis was used on the two subsets, observing changes in the main point of discussion the users had, signalling if changes brought to the app had influenced users and their opinions as a result. LDA topic modeling method was used along with descriptive analysis of textual data.