ANALISIS SENTIMEN MASYARAKAT TERHADAP KEPUASAN BELAJAR DALAM JARINGAN (DARING) DI MASA NEW NORMAL MENGGUNAKAN METODE NAÏVE BAYES CLASSIFIER
No Thumbnail Available
Date
2021-10
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
perpustakaan UR
Abstract
Online learning is a method of studying over the internet. Noawadays in Indonesia
began to implement the policy of online learning by utilizing learning management
system application as a learni006Eg medium,as one of the efforts to prevent the
widespread transmission of corona virus(Covid-19) in the world of education. This
policy of online learning activities remotely began to be widely discussed not only
through the real world but also on cyberspace, through social media such as twitter
so that it became a trending topic to be discussed. The purpose of thisstudy is to find
out the percentage of Pros and Contras on twitter’s users by using the one of the
classification method, Naïve Bayes Classifier. The data used for this study are
tweets that contains #belajardaring and #belajardarirumah from September to
October 2020 period. When collecting data using twitterscraper that is loaded
using pyhton. Then pre-processing the data that includes case folding, tokenizing,
normalization, stopword removal, and stemming. The labeling process resulting
two classes of data, positive and negative, with a total of 417 data. The results of
this study foundthat 69% of people's perception is Contras and the other 31% is
Pros, referring the online learning 2020 issues. Data that already has a class is
then divided (Split) by 70% training data and 30% testing data to be used in the
Naïve Bayes Classifier algorithm. The evaluation showed an accuracy score of
79.63%, a recallvalue of 88.49% and a precision value of 66.66% based on 417
tweets data,consisting 291 training data and 126 test data
Description
Keywords
Sentiment Analysis, ConfusionMatrix, Naive Bayes Classifier