WELFAKE – WORD EMBEDDING OVER LINGUISTIC FEATURES FOR FAKE NEWS DETECTION

Abstract

News is the only mode and set of information that helps the public to know what's happening everyday globally. We have started our path of reading news digitally, by which many "Fake news" are being circulated. Fake news is false or misleading information presented as news. Fake news often has the aim of damaging the reputation of a person or entity, or making money through advertising revenue. People unknowingly believe those fake news as original one without any analysis or study. Since the machine cannot read the words we use, we are going to use “ML model” to train our dataset to the machine. Our project is a two-phase benchmark model named WELFake based on word embedding where each and all words are converted into numerical values which is further processed to classify based on certain matching property using machine learning. The first phase preprocesses the data set and validates the veracity of news content by using linguistic features. The second phase merges the linguistic feature sets with WE(Word Embedding) and applies voting classification. The classification is based on words and meaning matching and this matching percentage should be above a threshold value we fix. In this paper we are going to discuss about choosing the best algorithm based on our needs and accuracy and complete the task successfully