measurement study of user feedback in mobile app stores

Open Access
Author:
Zeng, Ke
Graduate Program:
Computer Science
Degree:
Master of Science
Document Type:
Master Thesis
Date of Defense:
November 21, 2014
Committee Members:
  • Sencun Zhu, Thesis Advisor
Keywords:
  • mobile app stores
  • user rating and comments
  • review spam detection
  • user feedback analysis
  • topic model
Abstract:
User feedback is an important component for mobile app markets such as Apple App Store, because it is a crucial factor in determining popularity as well as downloads. Since positive ratings and reviews could attract more users and hence more profit, the rating system in mobile app stores has become the target of review promotion attackers. In our thesis, we consider three crucial issues about the user feedback of mobile app stores. First, we analyze characteristics of user feedback and develop statistical results. Second, we perform outlier analysis, especially of the users who give a large number of high ratings. These users would be the potential review promotion attackers. Third, we discover the main causes as to why people love or hate certain mobile apps by topic analysis. We apply our techniques to two user review datasets from Apple App Store and Amazon App Store. The results show that our techniques and analysis would be helpful to discover some characteristics of user feedback in mobile app markets.