The rapid spread of misinformation in the digital era poses significant threats to journalism, democracy, and society. TruthLens is a web-based fake news detection application designed to mitigate this issue. By leveraging Natural Language Processing (NLP) techniques and real-time internet scanning, TruthLens evaluates the authenticity of news headlines and articles. The system utilizes a PassiveAggressiveClassifier trained on large linguistic datasets and cross-references queries with live worldwide news via the Newsdata.io REST API. This dual-layered hybrid approach provides robust, high-accuracy classifications of text as either 'REAL' or 'FAKE,' equipping end-users with a reliable automated verification tool. Furthermore, the system is deployed as a responsive web application with a glassmorphism-styled frontend and a Python Flask REST backend, providing users with an intuitive, low-latency fake news detection experience. The offline machine learning classifier acts as a rapid evaluator for linguistic structure, while the online verification API cross-references real-time journalism databases to provide temporal context, successfully transcending the limitations of static NLP models.
Yash Kushwaha (Sun,) studied this question.