Mining, Correlation, and Management of Log Messages to Enhance The Internal Security Systems

A log message is a written message that records either events that occur in a hardware system, software runs, or a network connexion. Indeed, log messages are used to detect system anomalies, security vulnerabilities and cyber attacks. Ubitrak is a Montreal company specialized in the field of computer security. It offers various security services, including protection against attacks, testing and safety training payments. Ubitrak designed recently a system called UbiCloud, which helps prevent and detect malicious actions like phishing or unauthorized sending of confidential information based on log messages analysis. However, the rules that detect these malicious actions in UbiCloud are limited and inflexible. Ubitrak would abstract these detection rules in the form of a new methodology in order to make the detection rules more flexible and accurate. The purpose of this project is the specification of a methodology to detect anomalies and security vulnerabilities by mining log messages.

Faculty Supervisor:

Gabriela Nicolescu

Student:

Partner:

Ubitrak inc

Discipline:

Computer science

Sector:

Retail trade

University:

École Polytechnique de Montréal

Program: