ABSTRACT
Since the approach of the internet, many fake news and fabricated articles/contents observed widely. With the growing utilization of advancement and social media, buyers are making and sharing more information than some other time in recent memory. However, some individuals distributed counterfeit news with no significance to reality just to build the readership. Gossip distinguishing on social media is an essential issue. This paper talks about the methodology of machine learning and natural language processing to solve this problem. Use of TF-IDF (TermFrequencyInverse Document Frequency) and trained the data on four classifiers to explore which amongst them works well for this Indian dataset (https://github.com/Aks121/Fake-News-Analysis-on-Indian-Dataset ).The recall, precision and F1 scores help us figure out which model works best. The accuracy achieved so far is 95 on the ratio of 70:30 split dataset. The reason for this work is to approach the mechanized arrangement of the news stories utilizing machine learning. This can be used by the users to identify through the locales containing fake news. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.