An Analysis of Various Classification Algorithms for Fake News Detection

With the expansion of social media and advanced stages, the spread of fake news has ended up a noteworthy societal issue. This paper presents a comprehensive outline of machine learning strategies utilized for the detection of fake news. We talk about different approaches such as Natural Language Processing (NLP), unsupervised learning calculations, and machine learning models utilized within the recognizable proof of fake news. We dive into highlight designing, opinion investigation, and organize examination as key components of fake news detection frameworks. Moreover, we look at datasets, assessment measurements, and challenges related with detection of fake news. Through this analysis, we point to supply bits of knowledge into the current state-of-the-art techniques and headings for future investigate in combating fake news dispersal. This paper mainly used K-Nearest Neighbor, Jaccard similarity, LSTM, Decision Tree, and Logistic regression Classifiers for the classification. Main objective of this paper to verify the accuracy of models by performing the pre-processing the data and handle the imbalanced data more appropriately. The results of this papers show comparison of all classifiers with accuracy. Our results shows that KNN achieves 83%, 84% for Logistic Regression, 84% for Decision Tree, 85% for Jaccard Similarity & Random Forest is 85% and 98% for LSTM in our experiment.