Analisis Sentimen dan Pemodelan Topik Untuk Mengidentifikasi Topik Pandemi Covid-19 Pada Media Sosial Twitter menggunakan Naïve Bayes Classifier dan Latent Dirichleat Allocation

Herjuna Ardi Prakosa
Riyanto (Fakultas Sains dan Teknik Universitas Perwira)
Siti Nasiroh (Fakultas Sains dan Teknik Universitas Perwira)

JNANALOKA
Vol. 02 No. 02 September Tahun 2021

\The Corona virus or Covid-19 is of particular concern around the world. Many people talk about this virus through posting comments and opinions on Social Media. Twitter is one of the social media that is currently still widely used by the public to convey opinions in the form of a collection of words called tweets. Tweets related to the topic of Covid-19 can be classified using the Topic Modeling method to produce a data topic that is often discussed by Twitter users. One of the algorithms used to perform Topic Modeling is using Latent Dirichleat Allocation(LDA). In this study, LDA was used to find out what words appeared in the tweets about Covid-19 that the public had uploaded via Twitter. Before the tweet data is modeled with LDA, sentiment analysis is carried out first with the Naïve Bayes Classifier to produce Positive, Negative and Neutral sentiments.