NotiTrustAI – AI-Powered Fake Notification Analyze

About the project

Developed NotiTrustAI, an AI-powered NLP system designed to identify fake, manipulated, or AI-generated university notifications and distinguish them from genuine institutional notices. The project addresses the growing problem of fabricated academic announcements spreading through screenshots and social platforms.A custom labeled dataset was created from genuine institutional notifications and artificially generated manipulated counterparts. A complete NLP pipeline was then developed for text cleaning, normalization, tokenization, feature engineering, model training, and evaluation. Transformer-based text classification models were trained and validated against unseen data to assess their ability to recognize manipulated notification content.The project resulted in both a functional notification classification system and a reusable labeled dataset specifically created for this problem domain.