Electronic Thesis and Dissertation Repository

Thesis Format

Monograph

Degree

Doctor of Philosophy

Program

Computer Science

Supervisor

Kontogiannis, Kostas

Abstract

As Service-Oriented Systems rely for their operation on many different, and most often, distributed software components, a key issue that emerges is how one component can trust the services offered by another component. Here, the concept of trust is considered in the context of reputation systems and is viewed as a meta-requirement, that is, the level of belief a service requestor has that a service provider will provide the service in a way that meets the requestor’s expectations. We refer to the service offering components as service providers (SPs) and the service requesting components as service clients (SCs).

In this approach, we propose a technique that allows for evaluating and assigning trust to various service providers, by considering their ability to fulfill their clients’ expectations or policies, and assigning and updating the reputation of other service clients, based on their capabilities as recommenders for the aforementioned service providers. Service clients request opinions from other service clients (i.e. recommenders), when looking for an appropriate service to use. Different sources of recommendation are considered and opinions are transformed into evidences to be used by our proposed ranking algorithm. After the utilization of a service, the requesting client updates the trust and distrust values for said service, as well as the reputation values for the service's recommenders. In this work, service clients and service providers are considered software applications that coordinate with each other, and may include micro-services, software agents, smart contracts or any other distributed inter-networked resource, without making any assumptions as to what a service client or a service provider component is, as long as it is a component that issues or responds to requests.

The proposed approach has been shown, through implementing a prototype and executing appropriate experiments, to be very stable in the presence of high percentages of malicious users. Even when the dishonest clients account for up to 80%, honest users are able to receive accurate recommendations and select services that are able to fulfill their requirements. The proposed framework outperforms other approaches, in scenarios where malicious users are involved, by up to 20%, especially for higher percentages of users overvaluing or undervaluing services offered.

Our approach is, also, capable of quickly detecting deterioration of the QoS provided by a service provider, and supports the provision of incentives and compensations for allowing the selection of new and reformed services and ameliorating bad interactions caused by extenuating circumstances, respectively.

Summary for Lay Audience

As Service-Oriented Systems rely for their operation on many different, and most often, distributed software components, a key issue that emerges is how one component can trust the services offered by another component. Here, the concept of trust is considered in the context of reputation systems and is viewed as a meta-requirement, that is, the level of belief a service requestor has that a service provider will provide the service in a way that meets the requestor’s expectations. We refer to the service offering components as service providers (SPs) and the service requesting components as service clients (SCs).

In this approach, we propose a technique that allows for evaluating and assigning trust to various service providers, by considering their ability to fulfill their clients’ expectations or policies, and assigning and updating the reputation of other service clients, based on their capabilities as recommenders for the aforementioned service providers. In this work, service clients and service providers are considered software applications that coordinate with each other, and may include micro-services, software agents, smart contracts or any other distributed inter-networked resource, without making any assumptions as to what a service client or a service provider component is, as long as it is a component that issues or responds to requests.

The proposed approach has been shown, through implementing a prototype and executing appropriate experiments, to be very stable in the presence of high percentages of malicious users. Furthermore, the approach is capable of quickly detecting deterioration of the QoS provided by a service provider, and supports the provision of incentives and compensations for allowing the selection of new and reformed services and ameliorating bad interactions caused by extenuating circumstances, respectively.

Creative Commons License

Creative Commons Attribution-Noncommercial 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 4.0 License.

Download

Share

COinS