Trust Based Service Selection in Service Oriented Environment

Introduction

In recent years, the Web has started a steady evolution to become a distributed, dynamic, service-oriented environment (SOE) where applications can be automatically invoked by other clients over the Web. As an autonomous, platform–independent, and computational element, web services laid a foundation for constructing a distributed and collaborative web. A web service is a self-describing software system that can be advertised, located, and used across the Web based on a set of standards, such as WSDL, UDDI, and SOAP (Papazoglou, 2003). All these standards are based on Extensible Markup Language (XML), which allows Web services to interact with each other over the Web.

Although the development of web services seems very promising, there are some challenging problems. A plethora of Web services competing in offering similar functionalities bring a great challenge in selecting a suitable web service. The goal for service selection is to find a best set of services, available at runtime, reliable, taking service requesters preferences into consideration and so on. Thus, service requesters need to know not only the functional attributes, but also the non-functional attributes of web services, i.e., Quality of Service (QoS) properties. Web services are executed on the hardware of their respective providers, in different containers, separated by firewalls and other trust barriers (Sung, 2009). Therefore, the major challenge lies in providing a framework for evaluating the trustworthiness of service providers without any prior interactions. In an open environment, where anyone can publish services, a service requester may select a poor quality, time-consuming, expensive, or even harmful service. Therefore, mechanisms are needed to help service requesters to distinguish the quality of Web services. There is a growing consensus that the Web service ‘revolution’ would not eventuate until trust related issues are resolved (Briman, 2006).

Currently, researches on trust management in SOEs are gaining an increasing momentum. Most of the studies focus on establishing trust between entities based on third-party platform or intermediary mechanism (Singh, 2004, 2005; Maximilien, 2002). Research results also show that reliable reputation systems have helped to increase user’s trust in the Web (Dellarocas, 2003). Online businesses including eBay and Amazon have made a great profit by using reputation systems, where the clients provide feedbacks to the quality of service (QoS) after each transaction, and the mathematical methods are used to generate the reputation scores for service providers. Although such a reputation model provides a certain security measure for transactions, it also contains typical malicious behaviors, such as providing dishonest opinions and collusion (Sungkeun, 2005). The reputation-based model in Maximilien (2005) requires human participation in the Web services and the agents that recommend trust information are assumed to be the trusted parties without considering their possible malicious behaviors. In addition, these approaches seldom consider the social relations among entities and preferences of the service requesters However, in the real word we almost ask our trusted friends for their recommendations on products and services. Therefore, we make the following assumptions for the SOE based on the observations of the real world: