Game Development Reference
In-Depth Information
Figure 4. Affective Olympia architecture
student model comprises cognitive and affective
models , which are employed to infer the learner's
cognitive and emotional states. These models were
implemented using a PRMs approach and DBNs
as AI tools. Once the student model executes the
inference mechanisms, the results are forwarded to
the adaptable tutor model . The tutor model selects
the pedagogical, motivational or affective action
or actions that maximize the student's learning,
understanding or level of engagement. These ac-
tions are handled by the cognitive and motivational
modulators . According to the strategies selected
by the planner , the presentation content manager
module makes the necessary changes in the game
mechanics and the world model. The changes
made to the world model and the game mechanics
influence the behavior of the dynamic interactive
modules. The tutor model will be implemented
using Dynamic Decision Networks (DDNs), e.g.
Influence Diagrams, which enhance the potential
of DBNs through the incorporation of utility and
decision nodes (Jensen & Nielsen, 2007).
Olympia architecture is perceptual, intelligent,
adaptable and multimodal. Its perceptual feature
relates to the ability to infer the learner's cogni-
tive and emotional states during the learner's
interaction. It is intelligent, since it uses AI tools to
handle uncertainty involved in the teaching learn-
ing experience, e.g. DBNs and DDNs. Its adapt-
able feature is capable of adjusting pedagogical,
Search Nedrilad ::

Custom Search