A Risk and Similarity Aware Application Recommender System

Abstract

As mobile devices, especially smartphones, become more and more popular, the number of mobile applications increases dramatically. Though mobile applications provide users convenience and entertainment, they have potential threat to violate users’ privacy and security. In order to decrease the risk of violation, we propose a risk and similarity aware application recommender system, which recommends high quality applications to users. The system estimates applications’ risk based on the requested permissions and calculates the similarity between applications based on the ratings and the number of ratings. It recommends applications with the lowest risk and highest similarity based on a user’s current applications. The evaluation shows that the system works efficiently in recommending low-risk and high-similarity applications.