WIDIANTI, DITA (2025) ANALISIS SENTIMEN PADA X MENGENAI PELAYANAN PROVIDER TELKOMSEL MENGGUNAKAN METODE NAÏVE BAYES. Skripsi thesis, Universitas Teknologi Digital Indonesia.
![[img]](https://eprints.utdi.ac.id/style/images/fileicons/text.png) |
Text (HALAMAN DEPAN)
1_195410113_HALAMAN_DEPAN.pdf - Published Version Download (1MB) |
|
Text (BAB I)
2_19541113_BAB I.pdf - Published Version Download (450kB) |
|
Text (BAB II)
3_195410113_BAB_II.pdf - Published Version Download (161kB) |
|
Text (BAB III)
4_195410113_BAB_III.pdf - Published Version Restricted to Repository staff only Download (194kB) |
|
Text (BAB IV)
5_195410113_BAB_IV.pdf - Published Version Restricted to Repository staff only Download (578kB) |
|
Text (BAB V)
6_195410113_BAB_V.pdf - Published Version Download (14kB) |
|
Text (DAFTAR PUSTAKA)
7_195410113_DAFTAR_PUSTAKA.pdf - Published Version Download (14kB) |
|
Text (LAMPIRAN)
8_195410113_LAMPIRAN.pdf - Published Version Download (261kB) |
Abstract
This study analyzes public sentiment towards Telkomsel provider services expressed through social media platform X. Data was collected from January 1, 2022 to April 14, 2025. This study uses the Naïve Bayes Classifier method to classify public sentiment towards Telkomsel provider services and groups these sentiments into Positive, Negative, and Neutral categories. The stages carried out in this sentiment analysis study include data collection using tweet harvest, labeling using Laxicon Based, and Preprocessing which includes cleaning, normalization, case folding, tokenizing, stopword removal, and stemming. Tweet data is classified using the Naïve Bayes Classifier method with the Confusion Matrix model. The results of this study using 1296 tweet data that has gone through the labeling and preprocessing process show that 37% of X social media users commented positively, 34% commented neutrally, and 28% commented negatively. The application of the naïve bayes classifier method for sentiment classification obtained an accuracy of 52% with a ratio of 80:20. Testing with the Confusion Matrix model produces an average weighted average precision of 52%, recall 52%, and FI-Score 52%. Keywords: Sentiment Analysis. Naive Bayes Classifier, Laxicon Based, Telkomsel
| Item Type: | Thesis (Skripsi) |
|---|---|
| Additional Information: | Pembimbing: Danny Kriestanto, S.Kom., M.Eng. |
| Uncontrolled Keywords: | Sentiment Analysis. Naive Bayes Classifier, Laxicon Based, Telkomsel |
| Subjects: | A Karya Umum (General) > Ilmu Komputer (Computer Science) > Analisis Sentimen A Karya Umum (General) > Ilmu Komputer (Computer Science) > Metode Naive Bayes |
| Divisions: | Skripsi (S1) > Informatika (S1) |
| Depositing User: | Titis Pratiwi |
| Date Deposited: | 11 Agu 2025 01:34 |
| Last Modified: | 11 Agu 2025 01:34 |
| URI: | http://eprints.utdi.ac.id/id/eprint/10792 |
Actions (login required)
 |
View Item |
