An Intelligent Multi-Agent Recommender System for Human Capacity Building

摘要

This paper presents a Multi-Agent approach to the problem of recommendingtraining courses to engineering professionals. The recommendation system isbuilt as a proof of concept and limited to the electrical and mechanicalengineering disciplines. Through user modelling and data collection from asurvey, collaborative filtering recommendation is implemented using intelligentagents. The agents work together in recommending meaningful training coursesand updating the course information. The system uses a users profile andkeywords from courses to rank courses. A ranking accuracy for courses of 90% isachieved while flexibility is achieved using an agent that retrievesinformation autonomously using data mining techniques from websites. Thismanner of recommendation is scalable and adaptable. Further improvements can bemade using clustering and recording user feedback.