Text Mining Untuk Klasifikasi Emosi Pengguna Media Sosial Degan Algoritma Naïve Bayes

Hasnining, Ayu and Hazriani, Hazriani and Yuyun, Yuyun (2023) Text Mining Untuk Klasifikasi Emosi Pengguna Media Sosial Degan Algoritma Naïve Bayes. Patria Artha Technological Journal (PATJou), 7 (1). pp. 57-67. ISSN 2549614X

[img] Text (File Artikel)
671-1953-1-PB_Emosi-PATJou.pdf - Published Version

Download (738kB)
Official URL: http://ejournal.patria-artha.ac.id/index.php/patj/...


This research integrates data mining methods. The Naïve Bayes method is used to classify emotional classes on Twitter social media. Naive Bayes algorithm is a classification using probability and statistical methods. The background of this research, judging from the current situation of social media users, is more likely to express the emotions they feel in uploading social media statuses. By expressing frontal words. This must be a special attention by certain parties. In the research carried out by the author, it is expected to be able to find out the emotions of Twitter social media users by using the Naïve Bayes method. The author collects 5000 status data, which is then divided into two parts, namely 4000 status data as training data, and 1000 status data as testing data. The 4000 training data is processed by the system using the Naïve Bayes method so that it results in determining the emotional class as neutral 78.0%, afraid 1.0%, angry 2.0%, disgusted 2.0%, sad 5.0%, happy 8.0%, amazed 2.0%, bad 2.0%. Furthermore, the 1000 testing data is processed twice. The first, processed by the system using the Naïve Bayes method and resulted in the determination of the emotional class neutral 78.0%, fear 1.0%, angry 2.0%, disgusted 1.0%, sad 4.0%, happy 8.0%, amazed 2.0%, bad 4.0%. And the second, analyzed by the author and co-authors who are graduates of the psychology department and Indonesian language and resulted in happy emotions classes 28.4%, sad 19.4%, neutral 16.9%, bad 12.3%, angry 9.8%, disgusted 6.8%, amazed 3.1% and scared 3.0%. So it can be concluded that the determination of emotional class using Naïve Bayes, neutral emotion class has a higher presentation, while manual analysis of happy emotion class has a higher presentation.

Item Type: Article
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Depositing User: Dr. Eng. Hazriani S.Kom., M.T.
Date Deposited: 24 Sep 2022 02:46
Last Modified: 24 Sep 2022 02:46
URI: http://repo.handayani.ac.id/id/eprint/155

Actions (login required)

View Item View Item