Traditionally we got our day-to-day news from newspapers and trusted mainstream media channels that are required to follow strict codes of practice. Nowadays, online news has become a great source of information for netizens. People's news consumption habits have been modified as a result of the rapid growth and popularity of social media. The internet has introduced a new way of publishing and consuming news articles having very few editorial standards. However, most of this news are specious

more »

... nd deceptive that includes clickbait, propaganda, satire, and misleading headlines. Many news articles surfacing on social media are hoaxes and lookalike actual news, making it difficult for the netizens to distinguish between the two. Fake profiles are the major sources for disseminating such rigged news on social media. Finding these profiles has grown to be a crucial study area that has lately received a lot of attention. Detecting these fake news and fake profiles fast and effectively, an automated tool has become an essential requirement. In our research work, we have used 5 distinct machine learning models on 3 distinct datasets to diversify the results and then selected the best model with the highest accuracy for detecting fake articles and profiles disseminating such rigged news. We have also used TF-IDF for the tokenization of news articles. Distinct ML algorithms specifically Naive Bayes, Decision tree, Passive-aggressive, Support Vector Machine, and Logistic Regression are trained multiple times to achieve maximum accuracy for all the utilized datasets. The accuracy of all machine learning models is compared to find out which one is the most accurate and suitable for an assigned task. Then we put forward a novel model to check the authenticity of news articles and profiles using the same platform. Achieving better accuracy, we recommend our proposed model for verifying the legitimacy of articles and social media handles.