Abstract
Play Store and App Store have a large number of apps that are in competition as they share a fair amount of common features. User reviews of apps contain important information such as feature evaluation, bug report and feature request, which is useful input for the improvement of app quality. Automatic extraction and summarization of this information could offer app developers opportunities for understanding the strengths and weaknesses of their app and/or prioritizing the app features for the next release cycle. To support these goals, we developed the tool REVSUM which automatically identifies developer-relevant information from reviews, such as reported bugs or requested features. Then, app features are extracted automatically from these reviews using the recently proposed rule based approach SAFE. Finally, a summary is generated that supports the application of the following three use cases: (1) view users' sentiments about app features in competing apps, (2) detect which summarized app features were mentioned in bug related reviews, and (3) identify new app features requested by users.