Android application detection based on Apriori algorithm

In order to effectively detect malicious applications on the Android platform, it is proposed to use the Apriori algorithm to analyze the correlation between permission requirements among malicious programs from the perspective of permission patterns, in order to detect whether software in the Android application market is malicious. This algorithm requires the construction of frequent permission itemsets and permission relationship feature libraries, to improve detection speed, the permission relationship feature library can be clustered to remove redundancy and obtain a typical permission relationship feature library, which can then be used to detect malicious applications with a small number of permission features.This article analyzed 53081 applications from 5 domestic app stores. After detection by Apriori algorithm, 348 malicious programs were found, with a detection rate of 0.65%. The detection time for each application was approximately 2 milliseconds. Compared with other methods, this algorithm can quickly and effectively detect malicious programs, indicating that this method is more effective.