ANALISIS SENTIMEN TWITTER (X) TERHADAP UNDANG – UNDANG IBU KOTA NUSANTARA MENGGUNAKAN METODE NAÏVE BAYES DAN SUPPORT VECTOR MACHINE

Aminullah, Muhammad and Suhendro, Suhendro Yusuf Irianto (2024) ANALISIS SENTIMEN TWITTER (X) TERHADAP UNDANG – UNDANG IBU KOTA NUSANTARA MENGGUNAKAN METODE NAÏVE BAYES DAN SUPPORT VECTOR MACHINE. Masters thesis, Institut Informatika dan Bisnis Darmajaya.

[img] Text
COVER.pdf
Download (175kB)
[img] Text
ABSTRAK.pdf
Download (126kB)
[img] Text
Halaman Persetujuan dan Pengesahan.pdf
Download (5MB)
[img] Text
DAFTAR ISI.pdf
Download (107kB)
[img] Text
BAB I.pdf
Download (242kB)
[img] Text
BAB II.pdf
Download (336kB)
[img] Text
BAB III.pdf
Download (355kB)
[img] Text
BAB IV.pdf
Download (960kB)
[img] Text
BAB V.pdf
Download (83kB)
[img] Text
DAFTAR PUSTAKA.pdf
Download (182kB)
[img] Text
LAMPIRAN.pdf
Download (183kB)

Abstract

SENTIMENT ANALYSIS OF TWITTER (X) REGARDING THE NUSANTARA CAPITAL CITY LAW USING NAÏVE BAYES AND SUPPORT VECTOR MACHINE METHODS By: MUHAMMAD AMINULLAH 2221210023 The relocation of the capital city of Indonesia has elicited various responses from the Indonesian community, both positive, negative, and neutral. Many responses are conveyed through social media, especially Twitter, which is popular among Indonesians. By employing Support Vector Machine (SVM) and Naïve Bayes algorithms, this research conducted sentiment analysis and classified tweets into positive, negative, and neutral classes, using two language representation models, TF-IDF, and word embedding, and also employing various proportions of training and testing data such as 60/40, 70/30, and 80/20. The research result showed that SVM, especially with a linear kernel after adjusting hyperparameters, provided better performance, particularly at the proportion of training/testing data 80/20. Linear SVM (Hyperparameter Tuning) stands out as the best choice despite requiring longer training time. Additionally, the use of Word Embedding in modeling demonstrated better performance in terms of precision and recall compared to the TF-IDF method. This research contributes significantly to understanding the use of classification algorithms in the legislative process and provides a foundation for more effective decision-making in selecting suitable algorithms and modeling methods. Keywords: IKN, Naïve Bayes, Support Vector Machine, TF-IDF, Word Embedding, Hyperparameter Tuning.

Item Type: Thesis (Masters)
Subjects: Ilmu Komputer
eTheses
Divisions: Pasca Sarjana > Magister Teknik Informatika
Depositing User: Muhammad Aminullah
Date Deposited: 23 Aug 2024 03:16
Last Modified: 23 Aug 2024 03:16
URI: http://repo.darmajaya.ac.id/id/eprint/17075

Actions (login required)

View Item View Item