25.1 Introduction
Collaborative recommender systems are dependent on the goodwill of their users.
There is an implicit assumption – note the word “collaborative” – that users are in
some sense “on the same side”, and at the very least, that they will interact with
the system with the aim of getting good recommendations for themselves while
providing useful data for their neighbors. Herlocker et al. [10] use the analogy of
the “water-cooler chat”, whereby co-workers exchange tips and opinions.
However, as contemporary experience has shown, the Internet is not solely inhabited by goodatured collaborative types. Users will have a range of purposes