A multi-Agent framework for funded research supervision

Abstract

A supervisor's delayed intervention on a student's performance deficiencies seriously affects a research progress. Due to the demanding schedules of supervisors, it is a challenge for them to timely observe the performance deficiencies of the student. To mitigate this situation, we propose an agent-based system to enhance, control and manage the research supervision process. Specifically, the system is recommended for funded research projects as these research are subjected to penalties in case of failure. Four actors constitute the proposed system which are, a supervisor, a student, a system administrator and a software agent. Each actor has a role and is authorised to perform specific functions. An administrator can subscribe and unsubscribe members. Some functions of a supervisor include create and modify research milestones, verify a student's new task and view a student's progress. The system allows a student to view his/her progress, add a new task, upload a progress report and many other related functions. The main actor that facilitates the miscellaneous processes is the intelligent software agent. It manages the communication between a supervisor and a student, measures the performance and the completion, and tracks the different activities via alerting and reminding. In this paper, we highlight the pertinent issues in developing the framework. © 2015 IEEE.