混合特征的涉诈类APP分析模型的构建与研究

打开文本图片集
摘 要: 针对涉诈类APP分类难的问题,通过N-gram、主成分分析法等方法对Dalvik字节码与权限特征形成的特征码进行降维,利用K-Means++聚类算法构建APK涉诈家族分析模型,实现对涉诈类APP进行分类的目的。
关键词: 涉诈类APP; Dalvik字节码; 权限; K-Means++
中图分类号:TP391.4 文献标识码:A 文章编号:1006-8228(2023)12-81-04
Research on analysis model of fraud-related APP based on mixed features
Xia Yidan, Li Qiaoyu, Shi Junfan
(Zhejiang Police College, Hangzhou, Zhejiang 310000, China)
Abstract: Aiming at the problem of difficult classification of fraud-related APPs, N-gram, principal component analysis and other methods are used to reduce the dimension of the feature codes formed by Dalvik bytecode and permission features, and K-Means++ clustering algorithm is used to construct the analysis model of APK fraud-related family, so as to achieve the purpose of classifying fraud-related APPs.
Key words: fraud-related APP; Dalvik bytecode; permission; K-Means++
0 引言
随着智能手机在我国普及,人们上网的方式也随之发生变化。(剩余5573字)