Detecting source code similarity using low-level languages
Vedran Juričić · Information Technology Interfaces · 2011
The paper proposes a method for detecting source code plagiarism and calculating similarity between source files. Method does not compare original source code, but intermediate language code that is generated by compiler and converted to readable form by disassembler. The paper shows method's advantages and disadvantages and shows the results of its performance analysis that was performed on numerous test cases in different languages. Also, the paper proposes further steps for modifying comparison algorithm and improving its performance and precision.