Research on Text Classification for Identifying Fake News

Abstract

In the big data environment, massive news can always lead people to make their own judgments about events happening in society. The wrong guidance of fake news will lead to a negative effect on society. It is necessary to distinguish between real and fake news. In the traditional text categorization method, using the TF-IDF information of the words in the document as the weight matrix and applying it to the classifier. Inevitably, TF-IDF contains limited information, limiting the effect of classification. This paper proposed a method based on TF-IDF and Word2vec for identifying fake news, using SVM to verify its validity.