Philippine Twitter Sentiments during Covid-19 Pandemic using Multinomial Naïve-Bayes
John Pierre D
Abstract
News of the novel coronavirus (COVID-19) started circulating in the Philippines by the end of January 2020.Like any calamity or relevant news, people used social media platforms such as Twitter to voice their options.This paper examines the polarity of COVID -19 related opinions on Twitter from January to March 2020 by applying natural language processing.A total of 29,514 tweets were collected throughout the said dates, where 10% of which was manually labeled to train a Multinomial Naive Bayes classification model that achieved 72% accuracy.Results showed that 52% of the remaining tweets are positive, and 48% have negative sentiments.
Topics & Concepts
Coronavirus disease 2019 (COVID-19)PandemicBayes' theorem2019-20 coronavirus outbreakSevere acute respiratory syndrome coronavirus 2 (SARS-CoV-2)Multinomial distributionBayesian probabilityGeographyVirologyEconometricsComputer scienceArtificial intelligenceMathematicsMedicineInfectious disease (medical specialty)OutbreakDiseasePathologySentiment Analysis and Opinion MiningHate Speech and Cyberbullying DetectionDigital Marketing and Social Media