Grünewald, FrankaSiebert, MariaSchulze, AlexanderMeine, ChristophDesel, JörgHaake, Jörg M.Spannagel, Christian2017-09-292017-09-292012978-3-88579-601-5https://dl.gi.de/handle/20.500.12116/4789Parsing user access log files for retrieving additional information is a well known approach to obtaining additional knowledge of a web site. Most research interests focus on identifying users and tracking user behaviour. In contrast, this paper concentrates on the statistical evaluation of all available log data. Therefore, special items of a web page are detected, categorized and the access data of these items is analysed. This paper shows the related process using the example of an e-learning web portal. It starts with preparing the log data and analysing it manually. Afterwards the data is categorized according to the findings of the analysis and the results proven by manually selected test data. Based on categories, it is possible to recommend tele-teaching objects with similar access data and classify them automatically to these categories.enText/Conference Paper1617-5468