Using multi-agent system with publish/subscribe model in a mobile framework for capturing social network users´ behaviour

This paper focuses on an intelligent approach of using Smartphone to capture the users´ behaviour in a social network system, namely University-Cyber AgenT (UniCAT) system through a Publish/Subscribe Model (PSM) in a mobile network. The UniCAT is a multi-agent Social Network System (SNS) that captures and simulates the behaviours and daily activities of the people of Universiti Tunku Abdul Rahman (UTAR). The main purpose of the system is to capture the SNS users´ behaviours, thus reflected on software agents, with each agent structure built on a reasoning mechanism in Belief-Desire-Intention (BDI) model. A mobile version of UniCAT system is being proposed due to the convenience and feasibility in capturing more users´ behaviours frequently through wireless technology, though available in UTAR but scarcely enjoyed or functional. The uniqueness of this work is that users´ behaviours can be captured anywhere and anytime compared to existing desktop-based system and later information can be shared easily amongst UNICAT users because of the portability of mobile application, to save time, and for users´ convenience.