An effective API extraction scheme for dynamic binary similarity comparison

Abstract

Software piracy causes serious problems in many software-related industries. Dynamic binary similarity comparison methods can be used to detect software piracy. However, such methods generate large logs that require long periods of time to perform similarity comparisons. In this paper, we propose an extraction method that facilitates effective binary similarity comparisons. Using the cosine similarity and k-gram similarity, the proposed scheme is shown to be effective for binary similarity comparisons.