Software Watermarking Using Support Vector Machines

Software watermarking is a tool used to combat software piracy by embedding identifying information into a program. Most existing proposals for software watermarking have the shortcoming that they heavily rely on the stealth of watermark to prevent adversaries removing marks. Besides, the watermark is separate from the original program and can be destroyed via fairly straightforward semantics-preserving code transformations. In order to mitigate these problems, this paper introduces SVM-based watermarking, a non-stealthy approach to software watermarking based on the incomprehensibility of support vector machines(SVMs). The advantage of this technique is that software watermarking is handled as the knowledge embedded into SVMs and is closely associated with the program logic. Thus, it makes watermark more difficult to be destroyed and removed. The results of the experiment further indicate that the proposed method is a lightweight and effective software watermarking scheme.