SINAMBELA, VERONIKA PASKALIA (2024) Perbandingan Akurasi Metode Naive Bayes Classifier dan Support Vector Machine (SVM) pada Analisis Sentimen Ulasan Aplikasi MyTelkomsel. Undergraduate thesis, UPN Veteran Jawa Timur.
|
Text (cover)
cover.pdf Download (1MB) | Preview |
|
|
Text (BAB I)
BAB I.pdf Download (134kB) | Preview |
|
![[img]](http://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (BAB II)
BAB II.pdf Restricted to Repository staff only until 22 January 2026. Download (524kB) |
|
|
Text (BAB III)
BAB III.pdf Restricted to Repository staff only until 22 January 2026. Download (1MB) |
|
|
Text (BAB IV)
BAB IV.pdf Restricted to Repository staff only until 22 January 2026. Download (1MB) |
|
|
Text (BAB V)
BAB V.pdf Download (7kB) | Preview |
|
|
Text (DAFTAR PUSTAKA)
DAFTAR PUSTAKA.pdf Download (237kB) | Preview |
Abstract
MyTelkomsel is an application designed to help Telkomsel provider users carry out provider activities such as checking credit, checking quota, purchasing quota, checking quota active date and SIM card active period, and several other Telkomsel service features. Prospective users of the MyTelkomsel application can see reviews on Google Playstore and other users' experiences while using the application. Apart from that, these reviews can also be used as an evaluation for Telkomsel developers to improve aspects that are still not in accordance with the reviews. The Support Vector Machine (SVM) algorithm and Naive Bayes Classifier are two algorithms that can be used to classify these reviews. These two algorithms are two algorithms that are popularly used in classifying data in text form. This research makes a comparison between the SVM algorithm and the Naive Bayes Classifier in the MyTelkomsel Application review. The data taken is review data for one year, namely from 08 June 2022 to 07 June 2023, totaling 6019 reviews in Indonesian. In this sentiment analysis, sentiment is classified into three classes, positive, negative and neutral. Labeling is done with the Sentistrength_id lexicon. Feature extraction is carried out with TF-IDF word embedding. The methods used are SVM and Naive Bayes Classifier, which will then compare the results. Of the 6019 data obtained, there were 2699 negative sentiment data, 1870 neutral data, and 1450 positive data. After both algorithms were applied to three scenarios, SVM accuracy was superior with an average of 76% while Naive Bayes only reached 61%.
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | Q Science > QA Mathematics > QA76 Computer software T Technology > T Technology (General) |
||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Informatics | ||||||||||||
| Depositing User: | Veronika Paskalia Sinambela | ||||||||||||
| Date Deposited: | 22 Jan 2024 10:17 | ||||||||||||
| Last Modified: | 22 Jan 2024 10:17 | ||||||||||||
| URI: | http://repository.upnjatim.ac.id/id/eprint/20487 |
Actions (login required)
 |
View Item |