Detection of spam using email signatures

Abstract

Currently, unwanted emails are actively sent to the Internet. Millions copies of emails are sent simultaneously to various users. Often e-mails undergo minor modifications to complicate the detection of spam. The paper proposes options for determining the signature of e-mails that allow identify letters with the same content and structure. Content signature of the letter includes the basic phrases in the text of the e-mail with the exception of names, numeric codes, suspicious words that are not included in the dictionary. Structure signatures incorporate the same type of e-mails, such as paragraphs, tables, images. The paper shows the results of using signatures to detect e-mail spam.