User Concern in App Reviews, a study of perceived privacy violation among user sentiments and other contribute factors
Date
2022-05-02
Authors
Cheng, Yue
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Privacy, a significant factor in software usage, also provides software developers with additional insights into how applications can be improved. However, it is a delicate matter that peeks into user behaviour to the amount of information they are willing to share. With the rise of mobile applications, another concerning factor of user information collection also became prominent. The existence of user chatter on the Google app store can help identify whether privacy concern is problematic or not. However, little research has been conducted to study privacy violations and their contributing factors. In this project, we proposed using an LDA based privacy identification model that assesses the factors relating to user concerns with privacy matters on the Google App Store user reviews. A total of 45,114,727 rows of data were scraped from the Google play store, which were later filtered and processed into workable data. With the help of the Gensim LDA library, we can identify a coherence score of 0.604 and eight topics of various subjects. We later arranged these subjects into their corresponding categories, which could be used to analyze why specific privacy terms are more sensitive while others are not.
Description
Keywords
app review, privacy, machine learning