Abstract:

The presentation will attempt to cover three research themes,
united by their close relatedness to human aspects in computing. The first
part is related to believability of virtual environments and agents,
embodied in these environments. It includes an example of implementation
of awareness believability ­ the ability of such agents to relate to their
environment during interaction. This believability is formalised in terms
of three components - environment-, self- and interaction-awareness. The
second theme is related to the development of a visual language and
analysis system, which can inform us about the quality of interactions in
which we are engaged. It presents a way for encoding the way interactions
unfold, a method for developing visual languages and a visual analysis
system, geared towards patient-practitioner interactions in healthcare,
where patients usually have limited medical background. The third part
briefly presents work on including information about the structure of
relations between actors in predictive modelling cycle, aiming at
improving predictive accuracy.