Evidence Sets Approach for Web Service Fault Diagnosis

Abstract

The main objective of this paper is to diagnose the fault distribution of the web service process using expert subjective opinions. In order to capture the fuzziness, non-specificity and conflictive of human judgement, evidence set proposed by L. M. Rocha is utilized to represent the expert opinions linguistically. The fault propagation is illustrated via coherent fault tree, and then the minimal cut sets (MCS) are derived. The highest occurrence of MCS is determined with acceptable conflictive measure. Proper web performances' improvement can be done based on the criticality of each MCS.