Combination of web usage, content and structure information for diverse web mining applications in the tourism context and the context of users with disabilities

Abstract

This PhD focuses on the application of machine learning techniques for behaviourmodelling in different types of websites. Using data mining techniques two aspects whichare problematic and difficult to solve have been addressed: getting the system todynamically adapt to possible changes of user preferences, and to try to extract theinformation necessary to ensure the adaptation in a transparent manner for the users,without infringing on their privacy. The work in question combines information of differentnature such as usage information, content information and website structure and usesappropriate web mining techniques to extract as much knowledge as possible from thewebsites. The extracted knowledge is used for different purposes such as adaptingwebsites to the users through proposals of interesting links, so that the users can get therelevant information more easily and comfortably; for discovering interests or needs ofusers accessing the website and to inform the service providers about it; or detectingproblems during navigation.Systems have been successfully generated for two completely different fields: thefield of tourism, working with the website of bidasoa turismo (www.bidasoaturismo.com)and, the field of disabled people, working with discapnet website (www.discapnet.com)from ONCE/Tecnosite foundation.