3.2 Recommendation based on information seeking
process optimization
Figure 3 shows the concept of optimized process recommendation
model. The actual processes denote the processes
that were taken by users actually. When users seek
something from web pages, their actual process will be
recorded into logs. These logs are then used to extract the
information access behavior patterns and also used for
process mining. The aim of pattern extracting is to extract
information access behavior patterns of users from logs.
The aim of process mining is to extract successful seeking
experience and then optimize the seeking process from the
extracted successful seeking process according to the
similarity of users’ behavior patterns. Finally, the optimized
seeking processes are recommended to the target
users whose information access behavior patterns are
similar to the reference user group. Here, the reference user
conditions at the moment (i.e., reading a search result, page
up or page down). sequence, while the shorter-term process of actions consists
of a set of operations with a specific sequence. Generally,
in order to achieve a motive or a goal, users need to
complete a longer-term process of activity or a shorter-term
process of action. Therefore, based on these, we can infer a
motive or a goal through analyzing longer-term processes
of activities or shorter-term processes of actions