For Forensic Analysis of Fraud Detection in Digital Payments Using MCDM-TOPSIS Technique Under New <i>T</i> -Spherical Fuzzy Hypersoft Similarity Measures

In today’s financial systems, digital transactions are crucial for offering users significant convenience and rapid processing; however, as these networks develop, they have also become more sensitive to fraud, raising serious concerns for both consumers and financial institutions. To address these issues, particularly the presence of uncertainty and ambiguity in detecting fraud, advanced forensic analysis setups are needed for clear fraud detection and prevention. Employing the MCDM methods, our study aims to enhance fraud identification in digital transactions made through consumer devices. First, we propose some novel similarity measures for T-spherical fuzzy hypersoft set (TSFHSS) structure with their weighted version and similarity properties. The work combines TOPSIS techniques with new T-spherical fuzzy hypersoft similarity measures constructed using sine and tangent functions to enhance fraud detection accuracy and reliability. Furthermore, the study employs sensitivity analysis to determine the adaptability and predictability of the presented model across numerous conditions and decision-maker preferences, ensuring its applicability in real-world scenarios.