Abstract

Users vary in their interests, knowledge and experience. These differences result in different demands on an application. Applications that consider the personal needs of a user, e.g. present the content related to the user interests and adapt the visualization to the user preferences, will obtain a better user acceptance than applications not personalized. Since the user of a web application can easily switch to an application of another provider, the need of personalization is notably high. Users are represented by their corresponding user models. A user model is a representation of the user interests and his characteristics that can be used to personalize an application. The process of creating and updating theses models is called user modelling. The information that influences a model can be gathered explicitly from the user or implicitly. Implicit information can be determined by the process of web usage mining. An important part of the web usage mining is the user tracking. It gathers the behaviour of the user and formed the basis for the following analysis and personalization. To compare the user behaviour of several applications a uniform representation of the tracked data is needed. Furthermore a principle of the object oriented programming paradigm is to realize frequently required functionalities as reusable component. Within this diploma thesis a generic user tracking system for web applications will be developed. Since gathering user behaviour needs to regard the data privacy laws which are depict briefly. As the user tracking is intended to observe user behaviour, the definitions as well as the concepts of the behavioural science are introduced and described and how these concepts can be adopted to the tracking concept. In this context a set of behavioural pattern are chosen for being tracked by the system. In addition, the events are mentioned that imply a certain behavioural pattern. Also ideas and standards of semantic web are part of this thesis. The well known approach of Microformats is integrated in the user tracking system to retrieve information about the context which relates to the tracked user behaviour.