Exploring Man-Machine Interaction Issues using
Probabilistic Reasoning and Gaze Enhancement
KHIAT, Abdelaziz (0151206)
In the field of Man-Machine Interaction several approaches have been used to overcome
the inherent complexity of the task leading to amazing realizations and also
to new open problems.
In this work, we explore two interaction techniques. One of them is situated at
the decision making level while the other is at the user's behavior interpretation level.
From one side, we consider the problem of handling uncertainties in the act of
interaction using Bayesian Networks as a backbone technique for a database
From another side, we consider enhancing the human-computer interaction using
gaze information that provides a context for the proper action to take.
A proactive user support application has been realized based on this principle.
The application monitors the user in his task and proposes a service
when it is mostly needed.