Evaluating Quality of Web Services: a Risk-driven Approach

Composing existing web services to obtain new functionalities is important for e-business applications. Deficiencies of aggregated web services can be compensated involving a redundant number of them for critical tasks. Key steps lie in Quality of Service (QoS) evaluation and selection of web services with good quality in order to avoid frequent and severe faults of a composite service. This paper, first, surveys the existing approaches for QoS-driven web service selection. Then, it describes an improved selection algorithm that takes into account success rate, response time and execution cost of involved web services. Finally, we propose a novel approach for evaluating quality of redundant service compositions through analysis of risk related to the use of external web services.