A Framework for Capturing Distinguishing User Interaction Behaviors in
Novel Interfaces
By Samad Kardan, Department of Computer Science, UBC.
As novel interactive systems continue to be created, it is often difficult to
understand a priori which ensemble of user interaction behaviors are conducive
to good task performance. In this talk I'll describe our work on a user modeling
framework that relies on interaction logs to identify both classes of user
types, as well as the interaction behaviors characteristics of each type. This
information is then used to automatically identify the type/behaviors of new
users, with the long term goal of providing adaptive interaction support when
needed. I will also provide the results of a user study on the CSP applet from
the AIspace learning tools, demonstrating the effectiveness of this framework
for detecting different behaviour patterns.