Opinion Mining in Mental Health: Users’ (Many) Opinions at Your Fingertips

Smartphones and tablets are increasingly varied and accessible. Different manufacturers launch successive models at very different prices, disseminating technology and popularizing access to information. This fact, combined with the continuous evolution and expansion of the mobile internet infrastructure, has led to the proliferation of applications available in mobile stores. Companies worldwide develop new apps and make them available to hundreds of millions of users through app stores. The mobile application stores brought another phenomenon: a fast, accessible communication and feedback channel in the user’s hand. We have never received so many ratings, comments, criticisms, and praise that are publicly and instantly disseminated. The challenge became proportional to the amount of data available. The main objective of this research is to understand some of the reasons that encourage or hinder the adoption of technology by mobile application users. As a secondary objective, the research aims to investigate the possibility of enabling the automation of web scraping of comments and ratings in mobile applications in the Google Play store (Android system) and Apple Store (IOS system). For this experiment, applications focused on mental health were selected for web scraping. A Python script was developed to extract data from the mobile app stores. These data were later classified and analyzed. Preliminary results point toward a promising application of the technique, as it was possible to capture thousands of manifestations from users of a mental health application in an automated way. Mobile app stores offer a vast and unprecedented opportunity for data to improve the software development process. The use of an opinion mining technique demonstrated ample ability to enable the mapping of the user’s needs, frustrations, and opinions.