Sentiment-Aware Detection Method of Fake News Based on Linguistic Fuzzy Bi-LSTM

Fake news is a harmful fortification that can target the well beingness of individuals as well as organizations and make a spur about every misinformation. Fake news-based research grows in terms of contributions that apply deep learning, machine learning, and sentiment analysis, but there is an inadequate integration of fuzzy aspects with Bi-LSTM in language, which gives more accurate detection about news i.e, fake or real. In this work, we extended the classification model by combining BERT embedding with hedge based fuzzy logic with a deep learning Bi-LSTM algorithm to generate a more detailed linguistic based sentiment-aware model for fake news. On a dataset of nearly 12K texts labeled as real or fake and compared with other baseline algorithms that do not assimilate fuzzy based approach with misinformation detection, this research observed better results of 96.88% in terms of accuracy.