A model for filtering messages applied to an architecture to combat Phishing Scans and SPAMs

• Grant number: 08/09054-8
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: January 01, 2009
• End date: December 31, 2009
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computer Systems
• Principal Investigator: Adriano Mauro Cansian
• Grantee: Isabela Liane de Oliveira
• Host Institution: Instituto de Biociências, Letras e Ciências Exatas (IBILCE). Universidade Estadual Paulista (UNESP). Campus de São José do Rio Preto. São José do Rio Preto , SP, Brazil

Abstract

This project aims to show a form of mitigation of Phishing Scam and SPAM. This practice is very common in the daily life of electronic mail users as well as in the Internet communication infrastructure. So, it proposes a new model of message filtering using neural network and vectors of distancing of strings. The most used contention methodologies are analyzed, as well as their characteristics, advantages and disadvantages. A new model of message filtering, based on the collaboration of various methodologies and in the generation of signatures will be implemented. This model should be able to detect unwanted and fraudulent messages that can be shared between users of this system. Among the methodologies used, the analysis of frequency of characters, representation polynomial of frequency, functions of distribution of probabilities will be considered, and also the intersection of these. A neural network for the classification of messages will be developed, and for its training generated signatures are used. To analyze the messages of an MTA (Mail Transfer Agent) and to classify them in Phishing Scam or SPAM, vectors of distancing of strings and neural network will be used. A new architecture to combat Phishing Scans and SPAMs is proposed, aiming to contribute to the containment of attempted fraud by e-mail. This architecture shows the performance of the system coupled to an MTA, the stages of operation, and its collaborative use between various servers.