Résumé en anglais

Faced with b oth identity theft and the theft of means of authentication, users of digital services are starting to lo ok rather suspi- ciously at online systems. The b ehavior is made up of a series of observ- able actions of an Internet user and, taken as a whole, the most frequent of these actions amount to habit. Habit and reputation o er ways of recognizing the user. The intro duction of an implicit means of authenti- cation based up on the user's b ehavior allows web sites and businesses to rationalize the risks they take when authorizing access to critical func- tionalities. In this pap er, we prop ose a new mo del for implicit authen- tication of web users based on extraction of closed patterns. On a data set of web navigation connection logs of 3 , 000 users over a six-month p erio d we follow the exp erimental proto col describ ed in [1] to compute p erformance of our model