questions, the contributions and how these have been addressed in the papers,
the contextual description, the citations and educational use of papers leading to
this thesis, and finally the lessons learned from this work.
cyberspace services, e.g. mobile commerce and massively multiplayer online
game services? (Examples of customization include prediction of interesting
cyberspace service content and automatic provision of recommendations of
products and services)
conceptual software agent architecture for supporting general mobile commerce
services, and later peer-to-peer based distributed collaborative filtering for rec-
ommendations in such services. Scalable manner has been handled by adapting
and simulated empirical testing the architecture for a particular type of service
- massively multiplayer online games. Prediction behavior has been handled in
a scalable manner by empirical testing of the proposed incremental, decremental
and parallel classifier algorithms. Finally the empirical results from applying the
proposed incremental classifier applied on real web (from a web log) and synthetic
(from the developed MMOG simulator) datasets provide a closure.