A Supervised Machine Learning Approach to Identifying Online Fake News
Year: 2025
Type: article
Abstract: Fake news refers to intentionally and verifiably false stories that are largely disseminated through social media networks or the internet. Such news can be very persuasive, which makes it necessary to develop strategies to identify and critically assess news read and circulated on social media. This paper presents a model to detect online fake news using supervised machine learning algorithms. In this paper a fake_or_real_news dataset was used in feeding and training the machine learning algorithms. This dataset was preprocessed and extract using feature extraction. The dataset which have 4 columns originally was further divided into two columns which are the text and label columns. The label Columns was further processed to be a REAL column. In this paper we used three supervised machine learning algorithms in training our model. The three algorithms are as LogisticRegression, Support Vector Classifier, MultinomialNB. After training and evaluating the performance of the three models, the results shows that Logistic Regression had an accuracy of about 91.9%, Support Vector Classifier had an accuracy of about 86.8% and MultinomialNB had an accuracy result of 88.5%. Therefore, this paper recommends the use of Logistic Regression in detecting online fake news.
Author: Amadi, Chizoba Bright
Institution: Rivers State University
Language: English
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7104051381
FWCI:
Cites:
Cited by:
Related to: