Fact Checkr

Abstract

Here is the Abstract for your Final Year Project documentation, synthe sized from the information provided in your uploaded file. Abstract These are the Abstract of your Final Year Project documentation, which has been synthesised out of the information given in your uploaded file. Abstract In the age of digitalization, the spread of information over social media and online platforms is incredibly fast, and a very important issue arose: the spread of false information and fake news. The given project is the Fake News Detector, the all-encompassing web-based program that aims to automate the concept of evaluating the credibility of news articles by the means of the sophisticated multi-dimensional analysis. The system uses a hybrid scoring system, which combines four verification elements: machine learning based text reporting, source reputation reporting, clickbait scheme identifying and image integrity reporting based on perceptual hashing. The core classification engine uses a Supervised Logistic Regression model which is trained on a labeled set of 94, 408 articles using TF-IDF vectorization which identifies linguistic patterns related to deceptive articles. This machine learning based subdivision is assisted by a database of verified domains and a unique clickbaiting recognition model. The application is designed as a Micro services based system and uses react to entertain the user interface, node.js and express to do the back-end logic, and finally a Python flask service to do machine learning inferences. The results of the experiment prove that the system has a classification accuracy of 92.5 per cent in the test dataset. By providing features such as real-time URL scraping, live news exploration via NewsAPI, and explainable AI results, the Fake News Detector offers a scalable, accessible, and high performance solution for journalists, researchers, and the general public to combat misinformation effectively.