Abstract
Segregation of reliable and unreliable mobile applications and ranking the reliable Apps based on Ranking, rating and review based techniques. This leads to the importance of preventing ranking reliable application which has been widely recognized, and there is a limited understanding and research in this area. To this end, we propose a system for ranking reliable mobile Apps. To start with locate the ranking reliable Apps by mining the active periods, namely leading sessions, of mobile Apps. Such leading sessions can be control for detecting the local abnormality instead of global abnormality of App rankings. By a proof, we investigate this by three types of evidences, i.e., ranking, rating and feedback based reports, by investigating Apps' ranking, rating and review behaviors through various experimental, proved and statistical tests. In addition, we introduced an optimization based aggregation method to integrate all the evidences for detecting reliable Apps. Finally, we validate the effectiveness of the proposed system, and show the scalability of the detection algorithm.