3.2
Software Agents
17
mining in massively multiplayer online games, and proposes game usage
logs for the latter approach.
Web Agents covers agent technology, recommender systems and automatic e-
mail filtering. Two examples are:
1. Neural Networks for automatic e-mail and anti-spam filtering, Clark
2. How animations and software agent technology can be used to improve
This topic is relevant to the thesis since paper A and B utilize agent tech-
nology for wireless services, and massively multiplayer online games paper
C. Paper D presents empirical performance evaluation of the simulator pre-
sented in paper C.
Emerging Web Technology and Infrastructure covers grid computing, peer-
to-peer computing, soft computing (e.g. machine learning techniques for
classification, association rules and clustering) and web document prefetch-
ing. Two examples are:
1. Evolutionary algorithms for real-time prediction of user click-streams,
2. Classification of topic-specific web sites using Hidden Markov Tree
This topic is relevant to the thesis since paper F, G and H present approaches
for incrementally learning classifiers. The classification algorithm proposed in F
is empirically compared to existing classifiers in paper I. Paper B is related to
peer-to-peer computing.
In the next sections we first go into detail about software agents in mobile com-
merce and multiplayer online game settings (i.e. belonging to the web agents
topic), and second into classification based on soft computing (i.e. belonging to
the emerging web technology and infrastructure topic)
3.2
Software Agents
This section describes the foundation of software agents and then continues with
descriptions of software agents in mobile commerce and multiplayer online games.
The first part of this section is based on Tveit [2001a].