Evaluating the probabilistic side of proactive computing using cognitive modelling methodology and the statistical inference of user's cognitive states

摘要

In the following paper we present the work in progress, which aims to validate through theoretical and empirical study the probabilistic side of proactive computing. We choose the cognitive modelling methodology as an approach to implement the probabilistic inference of user's cognitive states during online activity. We create four cognitive models, which simulate various cognitive states of a user during the task of online search of medical terminology. Additionally, we elaborate several potential scenarios, which may take place during the given task, and we associate these scenarios with our models as the integral elements of the proactive system behaviour. Ultimately, by applying the rules of the Bayesian statistics, we envision to test the principles of proactive computing in the frame of probabilistic approach.