Fake News Detection and Classification Using Machine Learning

19 Nov

Authors: Sakshi Santosh Kasar, Dr. Anita Mahajan

Abstract: The rapid growth of online news, social networking platforms, and instant messaging has made information easier to access, but it has also increased the circulation of misleading claims. Manual verification alone cannot keep pace with the volume and speed of online content. This paper presents a machine-learning-based framework for detecting and classifying potentially fake news from textual content. The proposed workflow uses a labelled dataset, text cleaning, feature extraction, supervised classification, and evaluation through standard performance measures. CountVectorizer and Term Frequency–Inverse Document Frequency (TF-IDF) are considered for converting text into numerical features. Candidate classifiers include Support Vector Machine (SVM), Logistic Regression, Decision Tree, Random Forest, Naive Bayes, K-Nearest Neighbours (KNN), and XGBoost; an LSTM model may be explored where sufficient data and computing resources are available. The LIAR/LIAR-PLUS datasets are relevant options for experiments involving fact-checked political statements. Model quality should be assessed using accuracy, precision, recall, F1-score, and a confusion matrix rather than accuracy alone. As no verified experimental scores were supplied in the source documents, this paper describes a proposed framework and evaluation plan rather than claiming a measured winning model. The system is intended to assist users and fact-checkers, not replace human judgement or source verification.