Fake News Prediction Based On Natural Language Processing(nlp) And Machine Learning

Abstract

The Rapid Dissemination Of Misinformation Across Social And Digital Media Platforms Has Created A Growing Demand For Automated Mechanisms To Detect And Prevent The Spread Of Fake News. This Paper Presents A Machine Learning–based Fake News Prediction System That Utilizes Natural Language Processing (NLP) Techniques To Classify News Articles As Real Or Fake. The System Preprocesses Textual Content Using Tokenization, Stop-word Removal, And Lemmatization, And Transforms It Into Feature Vectors Through Term Frequency–Inverse Document Frequency (TF-IDF) Representation. A Passive Aggressive Classifier (PAC) Is Employed To Train And Predict Labels With High Efficiency And Low Computational Cost. The Proposed Approach Achieves Competitive Accuracy While Maintaining Interpretability And Scalability. A Lightweight Flask Web Interface Is Developed For Real-time User Interaction, Enabling Non-technical Users To Input Text And Instantly View Classification Results. Experimental Evaluation Demonstrates That The System Effectively Distinguishes False Information From Legitimate News, Contributing To The Reliability Of Online Information And Enhancing Trust In Digital Communication

Keywords