Hermansyah, Mas'ud (2022) Analisis Sentimen Twitter Untuk Mengetahui Kesan Masyarakat Tentang Pelaksanaan Pomprov Jawa Timur Tahun 2022 Dengan Perbandingan Metode Naïve Bayes Classifier Dan Decision Tree Berbasis Smote. Jurnal Informatika dan Teknologi Komputer (JITEK), 2 (3). pp. 249-255. ISSN 2809-9230

[img] Text (Turnitin Analisis Sentimen Twitter Untuk Mengetahui Kesan Masyarakat Tentang Pelaksanaan Pomprov Jawa Timur Tahun 2022 Dengan Perbandingan Metode Naïve Bayes Classifier Dan Decision Tree Berbasis Smote)
Analisis Sentimen Twitter Untuk Mengetahui Kesan Masyarakat Tentang Pelaksanaan Pomprov Jawa Timur Tahun 2022 Dengan Perbandinga.pdf
Download (1MB)

Abstract

Sentiment analysis is a method used to understand, extract, and automatically process text data to get the sentiment contained in an opinion. Sentiment analysis will be used to process comments made by the community or supporters of each participant of POMPROV East Java 2022 through various media, including Twitter, regarding the progress or results of POMPROV East Java 2022. The number of comments, the authors use data mining methods and algorithms to process the comment data to get information about the POMPROV East Java 2022 event. The Naïve Bayes Classifier and Decision Tree classification algorithms are used as tools to classify comments expressed by users. Based on the results of experiments that have been carried out four times according to the number of data splits and twice based on the algorithm used, it can be concluded that the use of the SMOTE algorithm can increase the accuracy of the various data split compositions used. The best results of the Naïve Bayes Classifier method are found in the 7:3 data distribution which increases the accuracy by 14.52% and the Decision Tree method in the 9:1 data division increases the accuracy by 9.45%.

Item Type: Article
Subjects: T Technology > T Technology (General)
Divisions: STI-S1
Depositing User: Mas'ud Hermansyah
Date Deposited: 18 Jan 2024 03:38
Last Modified: 18 Jan 2024 03:38
URI: http://repo.itsm.ac.id/id/eprint/1152

Actions (login required)

View Item View Item