A survey on trust and reputation models for Web services: Single, composite, and communities

• Post by: eargadmin
• 27 June 2021
• Comments off

Abstract

Web service selection constitutes nowadays a major challenge that is still attracting the research community to work on and investigate. The problem arises since decision makers (1) cannot blindly trust the service or its provider, and (2) ignore the environment within which the service is operating. The fact that no security mechanism is applicable in such a completely open environment, where identities can be easily generated and discarded makes social approaches such as trust and reputation models appealing to apply in the world of Web services. This survey classifies and compares the main findings that contributed in solving problems related to trust and reputation in the context of Web services. First, a high-level classification scheme partitions Web services into three main architectures: single, composite, and communities. Thereafter, a low-level classification within each architecture categorizes the trust and reputation models according to the technique used to build the trust value. Based on this classification, a profound analysis describing the advantages and shortcomings of each class of models is presented; leading to uncover possible topics that need further study and investigation. In particular, we discuss the challenging problem of having active malicious Web services in the composite and community-based architectures. Thus, the paper can be used by the future researchers as a roadmap to explore new trust and reputation models for Web services taking into account the shortcomings of the existing models.