A Markov decision-based meeting scheduling mechanism for Automatic Secretary System using Internet and Smart-agent Technologies (ASSIST)

This paper presents an automatic secretary system called ASSIST (Automatic Secretary System using Internet and Smart-agent Technologies), which provides basic secretarial services, particularly meeting scheduling. We formulate a meeting scheduling problem based on a Markov decision model. The backward induction algorithm is used to determine the optimal policy (i.e., the optimal number of time slots for scheduling purposes). The analytical results are presented to show the system´s behavior.