Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{195306,
        author = {K Varshitha and L Bala kishan and M Lasya and M Mukesh and Dr. S Shiva Prasad and Ms. U.Shireesha},
        title = {An Explainable Hybrid Fraud Detection Engine for Online Payments Using Rule-Based Analysis and Machine Learning},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {10},
        pages = {7876-7885},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=195306},
        abstract = {Online financial fraud is a problem because of the fast growth of digital payment systems. The old ways of finding fraud mostly use fixed rules, which do not work well against fraud patterns and often give a lot of false warnings. This project is about a way to find online payment fraud that uses a combination of rules and machine learning to get it right and do it quickly. The project uses something called SMOTE to make this way of finding fraud work better. This new way combines rules with machine learning to detect online payment fraud accurately and in real time which is a big improvement, over the old ways. Online financial fraud and online payment fraud are issues that need to be solved. The system deals with the issue of class imbalance by using a method called Synthetic Minority Over-sampling Technique or SMOTE for short. This helps the model learn about fraud patterns even when it only has a few examples of fraud to work with.A Random Forest classifier is used to figure out how likely a transaction is to be fraud. The system uses all of this information to come up with a fraud risk score that changes based on the situation. The system uses this fraud risk score to help figure out if a transaction's likely to be fraud. The Synthetic Minority Over-sampling Technique helps the model learn about fraud patterns and the Random Forest classifier helps to determine the fraud risk score. The proposed hybrid engine achieves significantly improved accuracy compared to conventional rule-based systems and provides explainable outputs to justify each prediction. Experimental results demonstrate that the system enhances detection performance, reduces false alarms, and is suitable for real-time fraud prevention in online payment platforms.},
        keywords = {Online Fraud Detection,Real-Time Transaction Monitoing, Hybrid Fraud Detection Model, SMOTE Data Balancing, Explainable AI(XAI), Risk Score Calculation, Random Forest Classifier},
        month = {March},
        }