A Review Paper on ThreatXplain: Explainable Artificial Intelligence For Malicious Threat Detection For Insider Theft And Phishing

Authors: Dhanashri Anil Aher, Kashish Namdev Badgujar, Jay Yogesh Chavan, Akash Bhausaheb Aher, Manjusha Gaikwad

Country: India

Full-text Research PDF File:   View   |   Download

Abstract: In today’s digital world, organizations face major cybersecurity threats such as data leakage and phishing attacks, which can result in loss of confidential information, financial damage, and harm to reputation. To address these challenges, this project proposes an intelligent system powered by Machine Learning that can detect and predict possible data leaks and identify phishing websites. The system works by creating and training a dataset using organizational log files. This helps it understand normal and abnormal data usage patterns. Registered users can access a dashboard where they can upload datasets or manually enter data parameters to check for signs of data leakage or authentication issues caused by employees. After analysis, the system generates a detailed report explaining how the leakage was detected and what factors influenced the prediction. Along with data leakage detection, the system also includes a phishing website detection module. In this module, users can enter any URL, and the system analyzes it using Machine Learning algorithms to determine whether the website is legitimate or a phishing attempt. The result includes an explanation that highlights the features and parameters that helped identify the phishing activity. Overall, this project aims to improve organizational cybersecurity by combining data leakage prediction and phishing detection in a single, user-friendly platform that helps organizations stay protected from internal and external digital threats.

Keywords: Explainable AI, SHAP, LIME, data leakage prediction, phishing detection.

Paper Id: 232838

Published On: 2025-12-04

Published In: Volume 13, Issue 6, November-December 2025

Share this