A Spam Filtering Model of Immune Based on Multi-Agent

摘要

According to the traditional spam filtering method effectively identify unknown characteristics and variability of the ability is not strong, according to the basic principle of biological immune system and multi agent technology proposed based on immune multi-agent spam filtering model. The model can realize the information exchange of each Agent, enhance the whole model "memory" mechanism, and effectively extract the information and variation characteristics of spam. Spam experimental simulation results show that the model and other models compared has better performance, can effectively improve the correct rate of spam model characteristics and reduce the false alarm rate.