摘要
为提高手机应用软件的安全性,提出一种基于Android系统的手机恶意软件检测模型;模型利用数据挖掘的方法对恶意软件中的敏感API调用进行数据挖掘,进而得到恶意软件检测规则;针对检测规则在检测非恶意软件时,产生较高误报率的问题,设计了加权FP-growth关联规则挖掘算法,算法在数据挖掘的两个步骤中,对敏感API调用加权,利用支持度阈值去除一些出现次数频繁而权重小的规则,降低了非恶意软件的误报率;实验结果表明,模型对恶意软件检测率达到81.7%,非恶意软件的检错率降低到11.3%。
In order to improve the security of mobile application software based on Android system , a mobile malware detection model is proposed. The model manipulates sensitive API call via data mining to obtain detection rules. To reduce the false positive rate when the rule is used to detect the non malware, a weighted FP--growth association rule mining algorithm is proposed. Based on weighting the sensitive API call, we employ a support threshold to eliminate the rules which preserve small weight and occur with high frequency. The experiments show that the model achevied a detection rate of 81.7%for malwares, and reduced the false positive rate to 11.3%.
出处
《计算机测量与控制》
2016年第1期156-158,共3页
Computer Measurement &Control
