| This thesis explores and experimentally examines the behaviour of mobile software agents (MAs) as they use various contribution strategies towards the provision of a public good. MAs are autonomous software objects that may start running on a host computer, stop what they are doing, move to another host computer, and start up from where they left off. Understanding the intricacies of MA behaviour is recognised as essential for further advancement and acceptance of MA technology.; The objectives of this research are to develop a greater understanding of MA behaviour and to further demonstrate the applicability of MAs to economic game theory modelling. Game theory studies multi-player games where there is strategic interaction between players that results in one or more outcomes. By combining the e-commerce and economics elements of the public good threshold contribution game together with MA technology an effective tool for the analysis of these elements is presented. Through use of this tool a large set of threshold contribution game experimental results have been produced.; From analysis of these results, observations are made that highlight the effectiveness of using MAs to represent players in an economic game. It is also observed that using an economic game to examine the behaviour of MAs provides a structured approach to establishing behaviour patterns of MAs.; The conclusions reached are that MAs are applicable for modelling economic games and that through modelling the public good threshold contribution game, relationships between information inputs, player strategies, and the resulting behaviour are more visible. |
