Penerapan Algoritma Support Vector Machine Analisis Sentimen Komentar Instagram (Studi Kasus: Komentar Instagram Najwa Shihab Vaksin Siapa Takut)

Desi, Nur Khasanah (2022) Penerapan Algoritma Support Vector Machine Analisis Sentimen Komentar Instagram (Studi Kasus: Komentar Instagram Najwa Shihab Vaksin Siapa Takut). Undergraduate Thesis thesis, Institut Teknologi Telkom Purwokerto.

	Text Cover.pdf Download (984kB)
	Text Abstrak (Eng).pdf Download (28kB)
	Text Abstrak (Indonesia).pdf Download (28kB)
	Text BAB I.pdf Download (175kB)
	Text BAB II.pdf Download (366kB)
	Text BAB III.pdf Download (655kB)
	Text BAB IV.pdf Restricted to Registered users only Download (447kB) | Request a copy
	Text BAB V.pdf Download (97kB)
	Text Daftar Pustaka.pdf Download (172kB)
	Text Lampiran.pdf Restricted to Registered users only Download (110kB) | Request a copy

Abstract

The covid-19 that struck at the end of 2019 caused many changes in everyday life. As a result, Indonesia through the Presidential Decree has implemented LargeScale Social Restrictions. One of the ways to do this is to get vaccinated against COVID-19. This information is disseminated by various media, one of which is social media Instagram. The well-known journalistic broadcaster Najwa Shihab also voiced his Covid-19 vaccination activities through his uploads of vaccines who are afraid. This upload garnered several responses, one of which was comments. From these comments, positive, negative, or neutral sentiments can be made. From this comment, it will be seen the value of positive, negative, and neutral sentiment using the Support Vector Machine or SVM algorithm as well as the accuracy of this SVM algorithm in classifying it. To see the performance of the SVM algorithm is evaluated using a confusion matrix. The dataset used in this study amounted to 1038 and obtained the highest accuracy results in the distribution of train data and test data of 70:30 in the testing use RBF kernel 72%, and the precision, recall, f-1 scores for negative comments were 74%, 85%, and 79%, neutral comments were 68%, 73%, and 71%, positive were 80%, 26%, and 39%. In future research, it is hoped that preprocessing can be done to be cleaner on the comment dataset used. Keywords Algorithm, Instagram, Sentimen, SVM

Actions (login required)

View Item