AnDarwin: Scalable Detection of Semantically Similar Android Applications

摘要: The popularity and utility of smartphones rely on their vibrant application markets; however, plagiarism threatens the long-term health these markets. We present a scalable approach to detecting similar Android apps based semantic information. implement our in tool called AnDarwin evaluate it 265,359 collected from 17 markets including Google Play numerous thirdparty In contrast earlier approaches, has four advantages: avoids comparing pairwise, thus greatly improving its scalability; analyzes only app code does not other information - such as app’s market, signature, or description increasing reliability; can detect both full partial similarity; automatically library remove similarity analysis. two use cases for AnDarwin: finding by different developers (“clones”) same developer (“rebranded”). ten hours, detected at least 4,295 that have been victims cloning 36,106 are rebranded. By analyzing clusters found AnDarwin, we 88 new variants malware identified 169 malicious differences requested permissions. Our evaluation demonstrates AnDarwin’s ability accurately large scale.