基于K-means算法的Android权限检测机制研究

Research on Android permission detection mechanism based on K-means algorithm

为了能够有效保护用户的个人隐私,设计了一种针对Android权限的检测机制。该机制采用静态分析技术研究不同类别应用程序的权限特征,首先根据权限的使用频度设置权限组,并借鉴TF-IDF思想为权限赋予权值;然后建立相应的数据库,计算应用程序的敏感值;最后使用K-means算法进行聚类分析,将应用程序进行分类。实验结果表明,该机制能够有效地检测出未知应用程序的危险程度。

In order to protect the users’privacy effectively,this paper designed a mechanism for Android permissions detection.This mechanism used static analysis technology to study the permission characteristics of different kinds of application programs.Firstly,it set up permission groups according to the usage frequency of permissions,and then assigned weights for permissions referring to the TF-IDF idea.Secondly,it established corresponding database to calculate the sensitive value of the application.Finally,it used the K-means algorithm to cluster analysis and classified the applications.Experiment results show that this mechanism can detect the degree of risk of unknown applications effectively.