Improvement and optimization of spam text filtering system
Ying Zhang, Xi Yang, Liu Yanqiu · 2012
By implementing feedback on English text judging process in anti-spam filtering system, a novel real-time variation keywords indexing anti-spam system is achieved to improve on the accuracy of spam filtering. Meanwhile the anti-spam system is optimized by genetic algorithm on feedback times, definition threshold and allocating weight along with weighted judgment on mail information such as subject etc. The simulation results show that better performance has been achieved in this feedback processing parameter-optimized anti-spam filtering system with as high as 98.2% average spam filtering accuracy.