ANALYSIS OF SENTIMENT OF INDONESIAN COMMUNITY ON METAVERSE USING SUPPORT VECTOR MACHINE ALGORITHM

Abstract

Metaverse was widely discussed right on October 28, 2021, when Mark Zuckerberg announced the company name change from Facebook to Meta. Regarding the sophisticated Metaverse concept, it is possible to make the pros and cons between the Indonesian people. This information is widely distributed in various media crews, one of which is Twitter social media. Tweets from social media can be used to obtain data and information as material for research on sentiment analysis. This study aims to determine the response of the Indonesian people to the Metaverse. Material data was obtained from Twitter tweets with keywords or Metaverse queries. Research conducted by the Support Vector Machine Algorithm in research with TF-IDF word weighting, the tests carried out with the results of the class division of positive sentiment 70.69%, neutral 15.85%, negative 13.46% will produce values ​​of accuracy, precision, recall, and f1-score. They use various data, namely, 90% training data and 10% test data. The result is high accuracy in testing many data, 2504 data, and the accuracy value is 81%, with an average value of 79% precision, 63% recall, and 57% f1-score.