E-Mail Spam Detection Based on Part of Speech Tagging

Ever since the emails became well-known tools in communication field, the problem of spams was associated with them. One of the most significant methods for filtering such junk email is diagnostic of those e-mails by applying some especial technics named as Data-Mining. In the presented paper, a new approach based on this strategy that how frequently words are repeated is proposed in which the key words in the evidence are found by usage of their repetition number (frequency). The key sentences, those with the key words, of the incoming e-mails have to be tagged and thereafter the grammatical roles of the entire words in the sentence need to be determined, finally they will be put together in a vector in order to indicate the similarity between the received emails. The proposed paper takes advantage of an extraordinary algorithm called K-Mean algorithm to classify the received e-mails. It is worthwhile to note that the so-called K-Mean algorithm follows some simple and understandable rules which are too easy to work with and this stands as a great privilege for this paper. The precision of the applied algorithm in diagnostic of the e-mails is 83 percent.

کلیدواژه ها:

K-Mean algorithm ، Spam e-mail ، data mining ، pos tagging ، vector model