Engaging the Dynamics of Trust in Computational Trust and Reputation Systems

Computational Trust and Reputation (CTR) systems are essential in electronic commerce to encourage interactions and suppress deceptive behaviours. This paper focus on comparing two different kinds of approaches to evaluate the trustworthiness of suppliers. One is based on calculating the weighted mean of past results. The second one applies basic properties of the dynamics of trust. Different scenarios are investigated, including a more problematic one that results from introducing newcomers during the simulation. Experimental results presented in this paper prove the benefits of engaging properties of the dynamics of trust in CRT systems, as it noticeably improves the process of business partners' selection and increases the utility.

• Type: Journal
• Year: 2010
• Where: KESAMSTA
• Authors: Agnieszka Danek, Joana Urbano, Ana Paula Rocha, Eugénio C. Oliveira
• Keywords: Computational Trust, Deceptive Behaviours, Intelligent Agents, KESAMSTA 2010, Trust