Proof of Clustering: an efficient and reliable blockchain-based clustering framework
Yamina Khenfouci, Yacine Challal · 2024
We propose a novel blockchain-based collaborative clustering framework, designed to enhance the efficiency and reliability of data clustering by decentralizing the clustering process across multiple nodes within the blockchain network. The system incorporates a consensus mechanism called Proof of Clustering to achieve the accuracy and consistency of clustering. This results thereby in ensuring confidence and reliability of the final clusters and prevents any breaches or corruption. To improve the performance of clustering process in the decentralized blockchain environment, data parallelism is employed. Through extensive experimentation and performance evaluation, we demonstrate the effectiveness and efficiency of proof of clustering consensus in various scenarios. Comparative studies with traditional centralized and distributed clustering methods showcase the superiority of our proposed solution by in terms of computational speed and resource utilization. PoC improves the Silhouette Score by 5.05% and reduces the Davies-Bouldin Index by 22.93%, indicating better cluster cohesion and separation in the PoC framework.