Sentiment Analysis of Online Lectures in Indonesia from Twitter Dataset Using InSet Lexicon

Musfiroh, Desi and khaira, ulfa and Utomo, Pradita Eko Prasetyo and suratno, tri (2021) Sentiment Analysis of Online Lectures in Indonesia from Twitter Dataset Using InSet Lexicon. MALCOM: Indonesian Journal of Machine Learning and Computer Science, 1 (1). pp. 24-33. ISSN 2222-2222

Text 11 - Published Version Download (3kB)

Abstract

The implementation of online lectures on various campuses in Indonesia has been emphasized since the outbreak of coronavirus. Online lectures are used as a solutionto continue teaching and learning activities during pandemic. But the implementation of online lectures raises a variety of opinions in the community, especially among lecturers. It also raises the attitude of the pros and cons fromvarious parties. For this purpose, data mining from Twitter analyzes sentiment on the topic of "online lectures". The data is classified into 3 classes, i.e. positive, negative, and neutral. This research was conducted with a lexicon-based approach technique using InSet Lexicon as an Indonesian opinion dictionary. The determination of the sentiment class for each sentence is obtained from the result of the polarity score calculation. Classification results from 5811 tweet data were found to contain 63.4% negative tweets, 27.6% positive tweets, and 8.9% neutral tweets. Testing of classification results was done by cross-validation method and confusion matrix witha comparison of training data and test data is 8:2gave accuracy value of 79.2%, precision by 72.9%, recall by 62.8%, and f-measure of 67.4%.

Type:	Article
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software
Divisions:	Fakultas Sains dan Teknologi > Sistem Informasi
Depositing User:	Utomo
Date Deposited:	23 Mar 2021 07:57
Last Modified:	23 Mar 2021 07:57
URI:	https://repository.unja.ac.id/id/eprint/16973

Actions (login required)

View Item