PENERAPAN ADAPTIVE SYNTHETIC MULTINOMIAL NAÏVE BAYES UNTUK MENGATASI IMBALANCED CLASS DATA PADA KLASIFIKASI SENTIMEN KENAIKAN BBM

ISMATULLAH, ISMATULLAH (2023) PENERAPAN ADAPTIVE SYNTHETIC MULTINOMIAL NAÏVE BAYES UNTUK MENGATASI IMBALANCED CLASS DATA PADA KLASIFIKASI SENTIMEN KENAIKAN BBM. Sarjana / Sarjana Terapan (S1/D4) thesis, Universitas Muhammadiyah Semarang.

Preview	Text 2. Halaman Depan.pdf Download (2MB) | Preview
Preview	Text 3. Abstrak.pdf Download (755kB) | Preview
Preview	Text 5. BAB I.pdf Download (714kB) | Preview
Preview	Text 6. BAB II.pdf Download (934kB) | Preview
	Text 7. BAB III.pdf Restricted to Repository staff only Download (2MB) | Request a copy
	Text 8. BAB IV.pdf Restricted to Repository staff only Download (4MB) | Request a copy
Preview	Text 10. Daftar Pustaka.pdf Download (722kB) | Preview
Preview	Text 9. BAB V.pdf Download (714kB) | Preview
	Text 12. Full Text.pdf Restricted to Repository staff only Download (3MB) | Request a copy
	Text 11. Lampiran-lampiran.pdf Restricted to Repository staff only Download (9MB) | Request a copy

Abstract

ABSTRACT Ismatullah, 2022, Application of Adaptive Synthetic Multinomial Naïve Bayes to Overcome Imbalanced Class Data in the Fuel Increase Sentiment Classification. Thesis, Statistics Study Program, University of Muhammadiyah Semarang. Supervisor: I. Fatkhurokhman Fauzi,M.Stat., II. Indah Manfaati Nur,S.Si.,M.Si The increase in the price of fuel oil (BBM) invites pros and cons in Indonesian society, so that many people give their sentiments about this Indonesian people's sentiment regarding the increase in fuel prices tends to be negative, giving rise to the problem of imbalanced class data. The ADASYN method is used to overcome this problem. This study aims to determine the public response in each province in Indonesia regarding the increase in fuel prices. Multinomial Naïve Bayes algorithm is used to classify text into positive, neutral, or negative sentiment categories. The results of this study indicate that regions with high information and communication technology development indexes such as DKI Jakarta, West Java and Bali tend to reject an increase in fuel prices. The majority of people in each province use the word "price" in giving their opinion regarding the increase in fuel prices, whether negative sentiment, neutral sentiment, or positive sentiment. The results of the average accuracy in all provinces in Indonesia are 0.882. These results indicate that the model that has been built using ADASYN is able to predict the testing data quite well. Keywords: Adaptive Synthetic, BBM, Multinomial Naïve Bayes, Sentiment Analysis.

Actions (login required)

View Item