User Modeling

All adaptive hypermedia systems (AHS) need to store some knowledge they have about each individual user, in order to use that knowledge to adapt the information to each user. Such a user model may contain information about very different aspects of a user, including: A common technique for storing all these kinds of aspects of a user is a set of attribute-value pairs. Each attribute can be seen as a dimension. The complete user model for a specific user forms a vector in the high-dimensional space formed by all attributes. If the "distances" along all dimensions are normalized (by transforming all possible data types that are possible to a fixed real interval, like [0..1], similarity between user models can be expressed using the distance between vectors in this high-dimensional space.