A structure free self-adaptive piecewise hashing algorithm for spam filtering

Nowadays, email spam problem continues growing drastically and many spam detection algorithms have been developed at the same time. However, there are several shortcomings shared by most of these algorithms. In order to solve these shortcomings, we present a structure free Self-adaptive piecewise hashing algorithm(SFSPH) together with its super method(SFSPH-S, which is much faster than SFSPH but has lower accuracy). Both of them are based on the extremum characteristic theory, robin fingerprint algorithm and optimization theory. Then we designed several experiments to evaluate the algorithms' performance, including accuracy, speed and robustness, by comparing them with the famous DSC algorithm and the Email Remove-duplicate Algorithm Based on SHA-1(ERABS). Our extensive experiments demonstrated the good performance and accuracy of our algorithm for spam filtering.