Effective preprocessing and feature analysis on Twitter data for Fake news detection using RWS algorithm

Abstract

The tremendous headway of web empowered gadgets develops the clients dependably strong in virtual redirection affiliations. Individuals from social affairs getting moment notices with respect to news, amusement, training, business, and different themes.  The development of artificial intelligence-based classification models plays an optimum role in making deeper analysis of text data. The massive growth of text-based communication impacts the social decisions also. People rely on news and updates coming over in social media and networking groups. Micro blogs such as tweeter, facebooks manipulate the news as faster as possible. The quality of classification of fake news and real news depends on the processing steps. The proposed articles focused on deriving a significant method for pre-processing the dataset and feature extraction of the unique data. Dataset is considered as the input data for analyzing the presence of fake news. The extraction of unique features from the data is implemented using Bags of relevant tags (BORT) extraction and Bags of relevant meta words (BORMW).