Puneet Modgil and M. Syamala Devi, Panjab University, India

Multiagent Systems are autonomous intelligent systems. In many academic institutions student admissions are performed after generating merit lists. Generation of merit lists is preceded by manual scrutiny of admission forms. This manual scrutiny is a knowledge-intensive, tedious and error-prone task. In this paper the design, implementation and testing of Multiagent System for Scrutiny of Admission Forms (MASAF) using Automatic Knowledge Capture is presented. MASAF consists of three agents namely: Form agent, Record agent, and Scrutiny agent. These three agents, using ontology, cooperatively fulfill the goal of highlighting the discrepancies in filled forms. MASAF has been tested by scrutinizing about 1000 forms and all of discrepancies found were correct as verified by human scrutinizer. Thus it can be concluded that using Multiagent system for scrutiny of forms can reduce human intervention, improve performance in terms of speed and accuracy. The system can be enhanced to automatically correct the discrepancies in forms.

Multiagent system, Knowledge Capturing, Scrutiny of admission forms, Ontology