A Gaming Framework for Modelling Competitive Service Industries

Candidate: Amit Kanwal
Type: Master of Science (MSc), School of Interactive Arts and Technology
Date: July 24, 2007
Senior Supervisor: Dr. Chris Shaw
Thesis: Download Thesis Document

Abstract

Due to the competitive nature of service industries, firms are often required to make sound business decisions in short periods.  Errors in marketing and operations strategies can result in loss of time and money.  Although computer simulation can aid in evaluating potential business models before they are deployed, the problem of making intelligent decisions becomes central to modelling rational behaviour of firms.

A multi-agent based gaming framework is proposed around a market model for service providers, where decisions as to how to allocate revenue are made using a multi-criteria optimization approach.  Kalman filtering is investigated as a means for estimating unknown parameters within the model, and basic consumer behaviour heuristics are implemented for reacting to market conditions.

The study demonstrates that although a more sophisticated business model implementation is necessary to exhibit realistic behaviour, based on initial evaluation, the framework comprising its core technologies is capable of facilitating such models.

Keywords: Multi-agent system; Multi-criteria optimization; Kalman filter; Gaming; Simulation.