A methodology for maintaining trust in virtual environments

Abstract

The increasing interest in carrying out business in virtual environments has resulted in much research and discussion of trust establishment between the entities involved. Researchers over the years have acknowledged that the success of any transaction or interaction via the virtual medium is determined by the trust level between trusting agent and trusted agent. Numerous publications have attempted to address the various challenges of assigning a trust level and building trust in an interacting party. However, the building and allocating a value of trust is neither easy nor quick. It involves high cost and effort. Hence, the ensuing research challenge is how to maintain the trust that has been established and assigned. Due to the dynamic nature of trust, the trust evolution, and the fragility of trust in virtual environments, one of the most pressing challenges facing the research community is how trust can be maintained over time. This thesis is an effort in that direction. Specifically, the objective of this thesis is to propose a methodology for trust maintenance in virtual environments which we term “Trust Maintenance Methodology” (TMM). The methodology comprises five frameworks that can be used to achieve the objective of trust maintenance.In order to achieve the aforesaid objective, this thesis proposes a: (a) Framework for third party agent selection, (b) Framework for Formalization and Negotiation of service requirements, (c) Framework for Proactive Continuous Performance Monitoring, (d) Framework for Incentive Mechanism, and (e) Framework for Trust Re-calibration.The framework for third party agent selection is used for choosing and selecting a neutral agent who will supervise the interaction between two parties. This is the first step of our methodology. The neutral agent is involved throughout the course of the interaction between two parties and takes a proactive-corrective role in continuous performance monitoring. Once both parties have chosen a neutral agent, they carry out a formalization and negotiation process of their service requirements using our proposed framework. This is in order to create an SLA which will guide the interaction between two parties. The framework for proactive continuous performance monitoring then can be used to evaluate the performance of both parties in delivering their service based on the SLA. If a performance gap occurs during the course of transaction, the third party agent will take action to help both parties close the performance gap in a timely manner. A key salient feature of our continuous performance monitoring is that it is proactive-corrective. Additionally, we design a framework for providing an incentive during the course of interaction to motivate both parties to perform as closely as possible to the terms of the mutual agreement or SLA. By the end of the interaction time space, both parties will be able to re-assess or re-calibrate their trust level using our proposed framework for trust re-calibration.Finally, in order to validate our proposed methodology, we engineered a multi-agent system to simulate the validity of the TMM. Numerous case studies are presented to elucidate the workings of our proposed methodology. Moreover, we run several experiments under various testing conditions including boundary conditions. The results of experiments show that our methodology is effective in assisting the parties to maintain their trust level in virtual environments.