Claroline is an open-source web-based application dedicated to learning and online collaborative work. The instance of Claroline at University of Namur (Webcampus) is the main communication channel between students and lecturers. Students may register to courses and download documents, receive announcements, submit their assignments, perform online exercises, etc.
We reverse enginneer the different behavioural models from the anonymized Apache access log provided by the IT support team of the university. This file (5,26 Go) contains all the HTTP requests made to access the Claroline instance from January 1st to October 1st 2013. We chose to consider only requests to access PHP pages accessible to users. It represents a total of 12.689.030 HTTP requests. We setup an automated way to infer the usage model from the log, as it would be intractable to do so manually. Each entry of the file represents a single access to a page: we had to reconstruct the behaviour of a user by grouping requests.
Please, see VIBeS: Claroline case study to download the inferred models.