Application of Blockchain-Artificial Intelligence towards Computerized Land Record Management System. *AI based Blockchain System for Sindh, Pakistan

Abstract

Traditional land registry systems have been sub- ject to duplicated entries, manipulation of land documents, and possible risk of legal dispute in urban areas of Sindh, Pakistan. Growing urban population demands preservation of land documentation with efficacy, reliability and transparency. Blockchain based systems have emerged in recent years as a promising solution to ensure immutable and duplication free on- chain storage of records in the form encrypted hashes. In this study, we have demonstrated the integration of blockchain and artificial intelligence for ensuring digitized land record system for province Sindh. We developed the experimental and analytical framework comprising an Ethereum-based smart contract and machine learning based verification mechanism to eradicate prevalent hindrances on land record management. We leverage experimental methodology with real Pakistani naming norms, geographical distributions across ten Sindh districts, utilized the dataset of 2000 residential records, having induced inconsistent records (duplicates: 2.0%, fraudulent entries: 1.85%, ambiguous data: 3.0%). AI layer application was accomplished through a random forest classifier pre-screened records with 95.2% accuracy, preventing dubious entries from being registered on the blockchain. The implementation ensures maintaining immutable records, successfully prevented the duplication and transparency was achieved with public availability of data records. With improved security, transparency, and fraud resistance, the system architecture provides a significant pathway for digitizing urban land records.