论文信息 - A Trainable Fuzzy Spam Detection System

A Trainable Fuzzy Spam Detection System

Electronic mail (e-mail) has been considered as one of the most convenient way to communicate among the users in the Internet. The rapid growth of users in the Internet and the abuse of e-mail by unsolicited users cause an exponential increase of e-mails in user mailboxes. Although there are several systems which use different AI techniques to filter out spam, there is hardly any system developed so far to filter e-mails using the fuzzy logic system. This paper presents the design and implementation of a trainable fuzzy logic based e-mail classification system that learns the most effective fuzzy rules during the training phase and then applies the fuzzy control model to classify unseen messages. Our findings imply that automatically trainable fuzzy spam filters are practically viable and can have a significant effect on spam detection.

[1] José María Gómez Hidalgo,et al. Combining Text and Heuristics for Cost-Sensitive Spam Filtering , 2000, CoNLL/LLL.

[2] Ray Hunt,et al. An analysis of the tools used for the generation and prevention of spam , 2004, Comput. Secur..

[3] Ian H. Witten,et al. Data mining: practical machine learning tools and techniques with Java implementations , 2002, SGMD.

[4] Harris Drucker,et al. Support vector machines for spam categorization , 1999, IEEE Trans. Neural Networks.

[5] Hideo Tanaka,et al. Trainable fuzzy classification systems based on fuzzy if-then rules , 1994, Proceedings of 1994 IEEE 3rd International Fuzzy Systems Conference.

[6] William W. Cohen. Learning Rules that Classify E-Mail , 1996 .

[7] Peter Norvig,et al. Artificial Intelligence: A Modern Approach , 1995 .

[8] Walter Daelemans,et al. TiMBL: Tilburg Memory-Based Learner, version 2.0, Reference guide , 1998 .

[9] Susan T. Dumais,et al. A Bayesian Approach to Filtering Junk E-Mail , 1998, AAAI 1998.