作者: Hardikkumar Rana , Mark Stamp
DOI: 10.1080/19393555.2014.975557
关键词: Source code 、 Theoretical computer science 、 Computer science 、 Malware 、 Detection rate 、 Context (language use) 、 Data mining 、 Function (engineering) 、 Variety (cybernetics) 、 Software 、 Code (cryptography)
摘要: ABSTRACTIn this article, we consider the problem of detecting software that has been pirated and modified. We analyze a variety detection techniques have previously studied in context malware detection. For each technique, empirically determine rate as function degree modification original code. show code must by substantially modified before fail to reliably distinguish it, our results offer significant improvement over previous related work. Our approach can be applied retroactively existing does not require access source code, hence it is both practical effective.