Modelling Evolving User behaviours
Friday 18 July 2008, 1300-1400
A talk by Jose Antonio Iglesias from the Carlos III University of Madrid (Spain).
Knowledge about computer users is very beneficial for assisting them, predicting their future actions or detecting masqueraders. Thus, to create and recognize the behaviour profile of a computer user is an interesting task.
However, a challenge in this task is how to accurately profile a user while his/her behaviour changes constantly. Thus, a user profile should be frequently revised to keep it up to date. In order to solve this problem, Evolving System can be very helpful.
This talk describes a way to create the model of computer user behaviour and classify it using Evolving Systems. In order to tackle this task, computer user behaviour is represented as the sequence of the commands s/he types. This sequence of commands is transformed into a distribution of relevant subsequences in order to find out a profile that defines its behaviour. Then, an evolving system is applied for classifying the profile.
Jose Antonio Iglesias is teaching assistant at the Carlos III University of Madrid (Spain). He received his Masterīs degree in Computer Science from Valladolid University (Spain). He has several papers about behaviour modelling and classification, especially in two different environments (RoboCup and UNIX User). Related to the RocoCup environment, he participated in the RoboCup Coach Competition 2006 held in Bremen (Germany). His research interests include evolving intelligent systems, modelling behaviour and their applications in real problems.