Abstract. Trust networks are directed weighted graphs whose nodes represent agents and edges represent trust between agents. This paper proposes a trust-based recommendation approach, which can recommend trustworthy agents to a requester in a trust network. We consider a good recommendation as one to an agent that the requester’s trusted neighbors trust highly. We relate the recommendation problem to the graph similarity problem, and define the similarity measurement as a mutually reinforcing relation. By calculating the vertex similarity between the trust network and a structure graph (a path graph of length three), we can produce a recommendation based on similarity scores that reflect both the link structure and the trust values on the edges.

We consider service-oriented computing (SOC) environments. Such environments are populated with services that stand proxy for a variety of information resources. A fundamental challenge in SOC is to select and compose services—to support specified user needs directly or by providing additional services. Existing approaches for service selection either fail to capture the dynamic relationships between services or assume that the environment is fully observable. In practical situations, however, consumers are often not aware of how the services are implemented. We propose two distributed trust-aware service selection approaches: one based on Bayesian networks and the other on a beta-mixture model. We experimentally validate our approach through a simulation study. Our results show that both approaches accurately punish and reward services in terms of the qualities they offer, and further that the approaches are effective despite incomplete observations regarding the services under consideration.

In service-oriented computing environments, computing resources are managed as services, which can be used directly or composed into larger services. Service-oriented architecture has been widely adopted in modern distributed environments such as for cloud computing. However, the problem of finding desired services has arisen. Finding desired services can be divided into two sub-problems: service discovery and service selection. The former one emphasizes how to find services that match consumers ’ requirements. The latter focuses on how to select best matched services. Service discovery usually finds services based on static functional attributes (e.g., service descriptions), whereas service selection tends to capture the dynamism of nonfunctional properties. For example, suppose a traveler is looking for flight tickets from Raleigh to Budapest. Service discovery

Abstract. Most of current service selection approaches in service-oriented environments fail to capture the dynamic relationships between services or assume the complete knowledge of service composition is known as a prior. In these cases, problems may arise when consumers are not aware of the underlying composition behind services. We propose a distributed trust-aware service selection model based on a Bayesian network for consumers to maintain their knowledge of the environment locally. Results show our model can punish and reward services in terms of QoS properties accurately with incomplete observations so that consumers can prevent themselves from interacting services with unsatisfying QoS. 1

Trust underlies effective interactions among autonomous parties. Ideally, a truster would base its trust in a trustee on the evidence of prior experiences with the trustee. But, such experiences arise between only a few trusters and trustees. Thus, a truster must rely upon the propagation of trust through a path of intermediary agents, each providing a trust assessment in the next. However, existing propagation approaches fail in settings such as social communities and product evaluations where a forward path isn’t available and we must consider a backward path from a rated entity to a rater. We propose Shin, a generalized trust propagation approach that incorporates a recent probabilistic method for revising trust estimates in trustees. Shin yields higher prediction accuracy than traditional approaches. In settings such as e-commerce and social networks, trust provides the basis for interaction among autonomous agents [1, 2, 3, 4, 5, 6]. We define a truster’s trust in a trustee as the truster’s belief that a future interaction with the trustee will yield expected outcomes [7]. Trust relationships naturally form a trust network, a weighted directed graph whose vertices represent agents and whose edges represent directed trust relationships, weighted with the level of trust. Each edge weight is affected by the outcomes of prior interactions and determines whether its source agent elects to pursue a future interaction with its target. A centralized reputation system consolidates the trust network. In a decentralized setting, each agent knows of only its own out-edges. A truster can reasonably decide whether to interact with a (prospective) trustee based on its estimation of the latter’s trustworthiness. A truster that lacks direct experience with a trustee would rely upon referrals leading to witnesses who provide testimony of their experiences with the trustee. Restricting the witnesses to direct experience helps avoid double counting evidence. Trust propagation [8, 9] means estimating trust over referral paths. Existing approaches consider only forward paths wherein each agent trusts the next. Our propagation technique, Shin (from the Chinese word for trust), applies even if no suitable forward path exists provided a path from both the truster and trustee to a common neighbor exists.

Abstract — Digital Business Ecosystem (DBE) is a decentralised environment where very small enterprises (VSEs) and small to medium sized enterprises (SMEs) interoperate by establishing collaborations with each other. Collaborations play a major role in the development of DBEs where it is often difficult to select partners, as they are most likely strangers. Even though trust forms the basis for collaboration decisions, trust and reputation information may not be available for each participant. Recommendations from other participants are therefore necessary to help with the selection process. Given the nature of DBEs, social network centrality measures that can influence power and control in the network need to be considered for DBE trust and reputation. A number of social network centralities, which influence reputation in social graphs have been studied in the past. This paper investigates an unexploited centrality measure, betweenness centrality, as a metric to be considered for trust and reputation.