A Fuzzy Logic Approach to a Hybrid Lexicon-Based Sentiment Analysis Detection Tool Using Healthcare Covid-19 News Articles

The delivery of unbiased news articles in the healthcare sector is one of the prominent problems in the fight against unvaccinated individuals as Covid-19 causes great skepticism among many groups of people. Companies such as Facebook have already integrated AI models for context to content that is rated by third-party fact-checkers to detect misinformation (“Fonctionnement du programme de vérification tierce de Facebook,” Fonctionnement du programme de vérification tierce de Facebook. https://www.facebook.com/journalismproject/programs/third-party-fact-checking/how-it-works?locale = fr_FR (accessed Sep. 06, 2021)). In this paper we use Natural Language Processing (NLP) and Sentiment Analysis to derive the content of news articles, an API is integrated to gather news articles from various sources using the newsapi (“News API – Search News and Blog Articles on the Web,” News API. https://newsapi.org (accessed Oct. 28, 2021)). Applying VADER, Text Blob, and Flair rule-based lexicons, we create a hybrid approach from the lexicons and combine each method, we then present a novel Fuzzy-Logic Lexicon Mamdani Rule-Base Multi Inference System (FLLMRBMIF) that can generate a final sentiment from each output of polarity, we then classify into a positive, neutral and negative result. The results demonstrate that it is possible to integrate a tool to classify the sources in real-time allowing more insightful information on biased news stories.