Agent Behaviour Monitoring in Virtual Organizations

Nowadays to capture an emerging opportunity, when beyond the abilities of single organizations, increasingly results the formation of a Virtual Organization (VO). Typically, the VO consists of a group of independent, autonomous, and heterogeneous entities that constitute agents agreeing to collaborate and jointly fulfill certain common goals. However, within the VO, agents expose different behaviours, and in case of frictions during agents' interaction and co-working, drastic results and even total failure of the VO may be followed. Monitoring the behaviour of both the individual agents and the sub-groups within the VO, can identify/predict problem areas and potential frictions among agents, what can then be used for enhancing the health of the collaboration, and thus the survival and success of the VOs. We propose a framework for controlling the VO partners' behaviours through monitoring their norms. The paper addresses three kinds of behavioural norms including the socio-legal norms, the functional norms, and the activity-related norms in VOs. The main norms ruling over socio-legal aspects applied to VOs are either clear and known, or defined within the VO's consortium agreement. Typically, during the VO creation phase, the high level agreed and expected agents' activities and inter-relations are represented through the VO contracts. Fulfillment of these coarse-grained activities corresponds to functional norms. Nevertheless, due to the VO's dynamic and evolutionary nature, during its operation phase many other agreements are continuously made among its involved agents. These agreements are then modelled as promises, while keeping these promises constitute the activity-related norms.