Application of Colored Petri Net in spam-filter rules property detection

Abstract

Currently, spam filtering methods and technologies appeared in large numbers. Among these methods, rule-based filtering which is widely applied on spam filtering system has high efficiency and strong-share. However, with the constantly updated empirical knowledge, the scale of the spam rule base has been increased rapidly which directly increases the difficulty to maintain, and also impacts the efficiency and the correctness of decision-making system. In this paper, the Colored Petri Net is applied to model and analyze the rule base, the modeling techniques and verficaiton techniques of abnormalities (including: redundant rules, conflict rules) detection are also introduced based on advanced nets system.