Evaluation and Validation of Trust using Fuzzy Inference Model for Agent Based Systems

It is compulsory for Agent based systems to interact with each other in order to achieve designated goals. Agents encompass several features i.e. competency, skills, achievement rate, trust etc. Trust is one of the important features for achieving the goal during communication. In real world, Trust is subjective in nature which is not useful in virtual communities. Virtual communities need precise value to trust on each other for realization of task. Fuzzy logic provides a way for definite conclusion for imprecise input information. We propose a system empowered with fuzzy logic to analyse trust. A fuzzy inference system is designed and implemented in MATLAB to include: (i) to reduce the subjectivity of trust; (ii) To help virtual community for interacting on the basis of trust. We also designed a graphical user interface to evaluate the trust. A validation is carried out by evaluating different agent based systems.