{"title":"AdDetect: Automated detection of Android ad libraries using semantic analysis","authors":"A. Narayanan, Lihui Chen, C. K. Chan","doi":"10.1109/ISSNIP.2014.6827639","DOIUrl":null,"url":null,"abstract":"Applications that run on mobile operating systems such as Android use in-app advertisement libraries for monetization. Recent research reveals that many ad libraries, including popular ones pose threats to user privacy. Some aggressive ad libraries involve in active privacy leaks with the intention of providing targeted ads. Few intrusive ad libraries are classified as adware by commercial mobile anti-virus apps. Despite such issues, semantic detection of ad libraries from Android apps remains an unsolved problem. To this end, we have proposed and developed the AdDetect framework to perform automatic semantic detection of in-app ad libraries using semantic analysis and machine learning. A module decoupling technique based on hierarchical clustering is used to identify and recover the primary and non-primary modules of apps. Each of these modules is then represented as vectors using semantic features. A SVM classifier trained with these feature vectors is used to detect ad libraries. We have conducted an experimental study on 300 apps spread across 15 categories obtained from the official market to verify the effectiveness of AdDetect. The simulation results are promising. AdDetect achieves 95.34% accurate detection of ad libraries with very less false positives. Further analysis reveals that the proposed detection mechanism is robust against common obfuscation techniques. Detailed analysis on the detection results and semantic characteristics of different families of ad libraries is also presented.","PeriodicalId":269784,"journal":{"name":"2014 IEEE Ninth International Conference on Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2014-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"50","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2014 IEEE Ninth International Conference on Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ISSNIP.2014.6827639","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 50
Abstract
Applications that run on mobile operating systems such as Android use in-app advertisement libraries for monetization. Recent research reveals that many ad libraries, including popular ones pose threats to user privacy. Some aggressive ad libraries involve in active privacy leaks with the intention of providing targeted ads. Few intrusive ad libraries are classified as adware by commercial mobile anti-virus apps. Despite such issues, semantic detection of ad libraries from Android apps remains an unsolved problem. To this end, we have proposed and developed the AdDetect framework to perform automatic semantic detection of in-app ad libraries using semantic analysis and machine learning. A module decoupling technique based on hierarchical clustering is used to identify and recover the primary and non-primary modules of apps. Each of these modules is then represented as vectors using semantic features. A SVM classifier trained with these feature vectors is used to detect ad libraries. We have conducted an experimental study on 300 apps spread across 15 categories obtained from the official market to verify the effectiveness of AdDetect. The simulation results are promising. AdDetect achieves 95.34% accurate detection of ad libraries with very less false positives. Further analysis reveals that the proposed detection mechanism is robust against common obfuscation techniques. Detailed analysis on the detection results and semantic characteristics of different families of ad libraries is also presented.