Abstract: Mobile App store ranking fraud is given the meaning that it refers to deceptive or malicious acts that perform the function of promoting the Apps in the popular list. Really, it grows more and more usual for developers of Apps to use dishonest actions, such as overstating sales of their Apps or providing untrue App ratings, to commit ranking fraud. While the importance of ranking fraud prevention has been comprehensively understood, research and knowledge on this front are limited. In this regard, in this paper, we discuss a thorough overview of ranking fraud and propose a ranking fraud detection system for mobile Apps. Specifically, we first propose to accurately identify the ranking fraud by mining active time, i.e., leading sessions, of mobile Apps. Such trailblazing sessions may be used for local anomaly identification instead of global anomaly of App rankings. Besides, we investigate three evidences, i.e., ranking based evidences, rating based evidences and review based evidences, by explaining Apps' ranking, rating and review behaviors through statistical hypotheses tests.

| DOI: 10.17148/IJIREEICE.2025.13362