ANALISIS SENTIMEN TERHADAP PELAKSANAAN PEMBATASAN SOSIAL BERSKALA BESAR (PSBB) PADA MASA NEW NORMAL MENGGUNAKAN METODE SUPPORT VECTOR MACHINE (SVM)

No Thumbnail Available

Date

2022-01

Journal Title

Journal ISSN

Volume Title

Publisher

Elfitra

Abstract

The spread of the COVID-19 virus has now become a pandemic because it has spread throughout the world, including Indonesia. The prositive case rate and the increasing death rate led the Indonesian government to make a regulation limiting community activities to break the chain of spread of COVID-19 in Indonesia. This regulation is a Large-Scale Social Restriction (PSBB. This becomes a polemic for some Indonesian people, especially those who do not have a fixed income. Many social media users discuss the implementation of PSBB in Indonesia. This study aims to find out how public’s sentiment towards the implementation of PSBB in Indonesia during the new normal period through tweets on social media Twitter using the Support Vector Machine (SVM) method. The data obtained for the implementation of the PSBB before the new normal was 322 tweet data that had been labeled while the tweet data for after the new normal was 653 data that had been labeled. The results obtained using the k-fold cross validation and confusion matrix method on the model created results in the highest accuracy rate is 81% with the distribution of data 90% : 10% for data before the new normal and for the data after the new normal, the highest accuracy value is 71% with the distribution of data 80% : 20%. The use of k-fold cross validation produces the highest accuracy value of 78% with fold value = 8 for data before new normal, while for data after new normal, the highest accuracy values is 73% with fold value = 1.

Description

Keywords

Confusion Matrix, K-Fold Cross Validation, PSBB, Sentiment Analysis, Support Vector Machine, Twitter

Citation

Perpustakaan

Collections