Trust in cloud environments is essential for enabling secure and reliable collaboration
among distributed users and services. Traditional models often
depend on centralized authorities, which can be inflexible and vulnerable in
dynamic cloud ecosystems. This thesis presents a decentralized approach to
trust management using Decentralized Identifiers (DIDs) and Verifiable Credentials
(VCs) to support privacy-preserving and adaptive reputation systems.
A prototype application is developed where users generate different
types of DIDs and anonymously vote for others based on trust, security, and
relevance. These votes are aggregated into VCs and used to calculate trust
scores through various methodologies, including direct, reputation-based,
and context-aware models. The system runs on a local blockchain and supports
use cases such as anonymous community or building-wide voting. Our
evaluation compares trust models based on accuracy, resistance to manipulation,
and score convergence, demonstrating the potential of decentralized
identity and trust frameworks to enhance digital collaboration in cloud environments.
|
