Unmasking_Misinformation_using_NLP_Techniques

The spread of misinformation and disinformation has become a growing concern in our digital age. In this project, we explore the use of Natural Language Processing (NLP) techniques to detect and unmask information pollution using the LIAR dataset. The task of detecting fake news is challenging and requires the use of different approaches such as machine learning and deep learning models. In this case, the liar dataset was used along with four different word embeddings (Tf-idf, glove, word2vec, and BERT) to train several machine learning models (Naive bayes, random forest, svm, logistic regression, and decision trees) and deep learning models (CNN, LSTM, and GRU). The results show that different embeddings and models led to different accuracy scores, indicating the importance of selecting the appropriate combination of embedding and model for fake news detection. Overall, it is essential to carefully consider the choice of model and embedding to achieve better performance in detecting fake news.