Sarcouncil Journal of Engineering and Computer Sciences

Abstract: The proliferation of digital payment systems has created unprecedented challenges in interpreting fragmented merchant identifiers within financial transaction data, necessitating sophisticated data science solutions for accurate merchant attribution and transaction enrichment. Contemporary financial institutions encounter substantial difficulties when processing abbreviated business names, inconsistent formatting, and cryptic merchant codes that obscure commercial entity identification in electronic payment records. Advanced computational techniques encompassing natural language processing, machine learning classification models, and deep learning architectures have emerged as critical enablers for transforming raw transaction data into standardized merchant profiles with comprehensive metadata. The multi-tiered framework incorporates sophisticated preprocessing protocols, entity extraction procedures, and comparative algorithmic approaches, including support vector machines, random forest implementations, and gradient boosting methodologies to achieve optimal brand prediction accuracy across heterogeneous transaction datasets. Implementation across diverse financial environments demonstrates significant improvements in customer segmentation capabilities, fraud detection mechanisms, portfolio management strategies, and regulatory compliance processes, while enabling enhanced market intelligence and business decision-making through enriched transaction analytics.