Android application classification and anomaly detection with graph-based permission patterns

摘要: Android is one of the mobile market leaders, offering more than a million applications on Google Play store. checks application for known malware, but abusively collecting users' data and requiring access to sensitive services not related functionalities are still present market. A permission system user-centric security solution against abusive malware that has been unsuccessful: users incapable understanding judging permissions required by each often ignore on-installation warnings. State-of-the-art shows current inappropriate end-users. However, lists do provide information about application's behavior may be suitable automatic analysis. Identifying key expected requests can help leverage abnormal simpler risk warning users. Applications with similar grouped into categories this work therefore analyzes category.In study, we propose methodology characterize normal category applications, highlighting requests. The co-required modeled as graph patterns central obtained using analysis metrics. evaluated performance classification allow choosing best metrics representing categories. Finally, study proposes privacy score threshold based efficiency proposed was tested set 9512 collected from malware. Display Omitted We build usage graph.We classify graph-analysis features.Among metrics, betweenness centrality weighted degree performed classification.We pattern-based metric applications.The showed high detection.