As the Internet grows in terms of interconnected computers, complexity and number of on-line services, computer security
becomes an even more relevant problem. On-line services may be requested to satisfy a number of 'constraints' such as
data privacy, service continuity, reliability and accountability. The more the strategic/economic relevance of on-line services,
the more the relevance of such requirements. On the other hand, security vulnerabilities on Internet services may be exploited
by criminal organizations to gain financial, political or military advantages.
- In such a context, it is fundamental to:
- analyze computer system events
- to detect violations of security policies, i.e., intrusions or intrusion attempts, and provide for automatic responses (Intrusion Detection)
- to find legal evidence in computers and digital storage media (Forensic Analysis)
- evaluate the effectiveness of the deployed security solutions (Evaluation)
Unfortunately, these problems are addressed only partially by current security solutions. The application of pattern recognition
and machine learning techniques can help improving state-of-the-art solutions, towards the above goals.
The "Computer Security" Technical Committee (TC) of GIRPR (Italian Group of Pattern Recognition Researchers)
is composed by researchers and pratictioners who adopt pattern recognition techniques to develop computer security solutions.
The goal of this technical committee is to build a meeting point for information exchange and brainstorming to support
the development of innovative solutions for coping with current security threats.
Department of Electrical and Electronic Engineering
University of Cagliari, Italy