Design and Evaluation of a Fintech-Enabled Personal Loan Decision Support System

Thesis Abstract

The increasing digitization of financial services and the rapid growth of fintech enterprises have significantly transformed the landscape of personal lending, necessitating the development of more accurate, efficient, and user-centric decision-making tools. Despite this progress, many financial institutions continue to rely on traditional credit assessment methods that often lack real-time responsiveness and may not fully leverage advanced data analytics or machine learning techniques. This study aims to design and evaluate a fintech-enabled personal loan decision support system (PLDSS) that enhances credit risk assessment accuracy, improves decision-making efficiency, and positively impacts borrower satisfaction. The specific objectives are to identify key factors influencing personal loan approval decisions, develop an integrated decision support prototype leveraging machine learning algorithms, and empirically evaluate the system's effectiveness through user testing and performance analysis. The research adopts a mixed-methods approach, combining quantitative and qualitative techniques to ensure comprehensive insights. The quantitative component involves a quasi-experimental design with a sample size of 300 loan officers and 1,200 loan applicants from two major banking institutions, selected through stratified random sampling. Data collection instruments include structured questionnaires for loan officers, semi-structured interviews for stakeholders, and system logs capturing user interactions and decision outcomes. Quantitative data will be subjected to multiple regression analysis and receiver operating characteristic (ROC) curve evaluation to measure the predictive accuracy and discriminative power of the developed system. Qualitative data from interviews will undergo thematic analysis to explore user perceptions, usability issues, and contextual factors affecting adoption. The study hypothesizes that the fintech-enabled decision support system will deliver statistically significant improvements in the accuracy of credit risk predictions, reduce loan processing times, and elevate user satisfaction levels compared to traditional methods. The anticipated findings are that machine learning models such as Random Forests and Gradient Boosting Machines will outperform conventional credit scoring models in predictive performance, with an accuracy increase of approximately 12%. Additionally, the system is expected to decrease decision turnaround times by over 25%, and qualitative analysis will reveal enhanced user confidence and trust in automated decision-making processes, albeit with some reservations regarding algorithmic transparency. This research contributes to the existing body of knowledge by demonstrating how advanced data analytics within a fintech-driven framework can optimize credit decision processes and inform best practices for integrating digital tools into banking operations. The study extends theoretical understanding through the application of the Theory of Planned Behavior and the Technology Acceptance Model (TAM), elucidating factors that influence the acceptance and sustained use of decision support systems in financial contexts. In conclusion, the study underscores the potential of fintech innovations to refine personal loan approval procedures, mitigate credit risk, and enhance customer service experience. Based on the findings, it recommends that financial institutions invest in developing customizable, transparent, and user-friendly decision support systems powered by machine learning. Additionally, policies should encourage ongoing training for loan officers to interpret algorithmic outputs effectively and foster customer trust. Future research avenues include exploring the system's scalability across different banking products and assessing long-term impacts on default rates and financial inclusion. Overall, the study provides a vital blueprint for banking stakeholders aiming to harness digital technologies for superior credit risk management and operational excellence.