Olariu, StephanMukkamala, RaviAljohani, Meshari2024-07-312024-07-312024-07https://hdl.handle.net/20.500.14154/72749Centralized marketplaces provide reliable reputation services through a central authority, but this raises concerns about single points of failure, user privacy, and data security. Decentralized marketplaces have emerged to address these issues by enhancing user privacy and transparency and eliminating single points of failure. However, decentralized marketplaces face the challenge of maintaining user trust without a centralized authority. Current blockchain-based marketplaces rely on subjective buyer feedback. Additionally, the transparency in these systems can deter honest reviews due to fear of seller retaliation. To address these issues, we propose a trust and reputation system using blockchain and smart contracts. Our system replaces unreliable buyer feedback with objective transaction assessments. Performance challenges of blockchain-based systems are tackled through three innovative schemes, resulting in a substantial improvement over the baseline approach. Furthermore, we proposed a decentralized marketplace utilizing blockchain-based smart contracts to address privacy concerns in buyer reviews that arise from the transparency of decentralized marketplaces. This enables buyers to use one-time identities for reviews to promote anonymity. This system ensures that buyers provide reviews by requiring a review fee, which is fully refunded after the review is submitted. Moreover, we proposed a trust and reputation service based on Laplace’s Law of Succession, where trust in a seller is defined as the subjective probability that they will fulfill their contractual obligations in the next transaction. This method accommodates multi-segment marketplaces and time-varying seller performance, predicts trust and reputation far into the future, and discounts older reputation scores. In addition, we propose SmartReview, an automated review system utilizing blockchain smart contracts to generate objective, bias-free reviews. The review module is designed as a smart contract that takes the contract terms and the evidence provided by the buyer and seller as inputs. It employs advanced computer vision and machine learning techniques to produce quantitative and qualitative reviews for each transaction, ensuring objectivity and eliminating reviewer bias. Lastly, we introduce a structured blockchain architecture featuring a layered approach. This architecture includes mechanisms for secure transaction recording and efficient query retrieval through auxiliary indexing, demonstrating significant advancements in decentralized data management.170en-USBlockchainSmart contractReputationTrustDecentralized marketplacePrivacyThesis