计算机科学
Android(操作系统)
指纹(计算)
匹配(统计)
假阳性率
数据挖掘
节点(物理)
人工智能
模式识别(心理学)
工程类
数学
统计
结构工程
操作系统
作者
Tengfei Tu,Hua Zhang,Yangye Hu,Dong Cui,X. Y. Zhai
标识
DOI:10.1016/j.jisa.2023.103527
摘要
Some existing methods for detecting repackaged apps use layout files to generate a global layout tree, but this can result in node positions affecting feature accuracy and repeated layout matching affecting model accuracy. In our paper, we propose ReHunter, a detection method for repackaged apps based on layout fingerprint. ReHunter uses a weight-based layout similarity algorithm to eliminate the influence of node position on features, and applies rules to prevent layout matching between applications, addressing the detection interference caused by repeated layout matching. Our experiments demonstrate that ReHunter achieves higher accuracy, with a false positive rate of 1.8% and a false negative rate of 1.7%.
科研通智能强力驱动
Strongly Powered by AbleSci AI