Prosiding Seminar Nasional Teknoka
Vol 8 (2023): Proceeding of TEKNOKA National Seminar - 8

Analisis Sentimen Masyarakat Terhadap Mainan Anak “Lato-Lato” Menggunakan Metode Support Vector Machine (Svm) Pada Media Sosial Youtube

Ahmad Roshid (Unknown)
Erizal, S.Kom., M.Kom. (Unknown)
Rahmi Imada, S.Kom., M.Kom. (Unknown)



Article Info

Publish Date
13 Dec 2023

Abstract

Lato-lato is a children's toy that has gone viral in Indonesia. This children's toy is quite controversial in Indonesian society because this toy creates quite a noise when played but also has several positive impacts for children. The purpose of this research is to analyze the sentiment of Indonesian people towards lato-lato toys from video comments on YouTube social media. The processes carried out are data crawling, preprocessing, normalization, labeling, and SVM classification. After the SVM classification process with a linear kernel, an accuracy of 81% was obtained. For positive sentiment, the precision is 80%, recall 85% and f1-score 82%, for negative sentiment, the precision is 83%, recall 77% and f1-score 80%. From the results of the sentiment analysis, more Indonesians respond positively to lato-lato, which indicates that Indonesians see lato-lato as not a problem and there are many positive benefits from lato-lato.

Copyrights © 2023