Abstract: This project's purpose is to use Natural Language Processing methods to identify misleading and news reports that originate from unreliable sources. only a count vector (Term Frequency Inverse Document Frequency) (word sizes compared to how (word sizes related to how often they are used in other articles in your dataset) was produced using values from this capacity vector and those features with the relevant point of 1.4 and that feature with relevant point 2.0 (or equal importance) (word) The semantic models, however, neglect aspects such as word classification and definition. Two documents with absolutely various word counts can refer to the equal thing. As a result, the data science area has put in place different actions to determine this problem. Facebook is engaging in a challenge on Kaggle to separate fabricated report stories from feeds on their social network using AI. opposing false news is an honest job Can you separate real and news from fakes? Thus, the proposed research would have the incorrect and the actual news datasets as knowledge and use the NB classifier to build a standard that matches articles by the words they introduce. owing to the extended number of online information references, it is tricky to know what is valid and what is unreliable Therefore, the issue of "fake news" has increased further publicity. This research looks at traditional and up-to-date approaches for defining accuracy and falsity in text format, as well as how and why it happens. This paper links Nave Bayes Classifier, support vector machines, and semantic analysis to recognize fake news, getting up with a system