A Blockchain Based Decentraized Search Engine Framework with Hybrid Cosine Similarity

Abstract

A blockchain based decentralised search engine to improve conventional search engines for retrieving and storing data with the help of IPFS, Elasticsearch, TF-IDF-CF ranking algorithm and blockchain technology. It is also improving the transparency, security, and efficient search result.  This work permits the analysis of user queries and improves the online visibility for keyword searches. The rapid growth of online information has made search engine optimisation for digital content. A traditional search engine has centralised control, limited data transparency, privacy, and ranking methods. Blockchain is not only for cryptocurrency-related information. It focuses on the web search platform and any related infrastructure, where the control is not centralised. This work aims to address these limitations by designing a data integrity, transparency, and manipulation ranking algorithm for web information retrieval. The proposed system has to create a tamper-resistant, verifiable and managed search index in this environment. To enhance the searching techniques, this framework integrates TF-IDF, Collection Frequency and cosine similarity to improve the ranking method for a   particular user’s query. Blockchain storage is integrated with IPFS to improve fault tolerance, while smart contract facilities secure validation and automated trust management. It achieved 91% precision, 88% Recall and 89% F1-Score through blockchain based search engine techniques.  This research underscores the potential of blockchain to redefine search engine architecture by promoting privacy, transparency, and user control, paving the way for more reliable and trustworthy digital information retrieval systems. Future work will focus on Layer-2 Blockchain solutions for enhancing scalability, AI based ranking techniques and large real-world dataset for further search performance.