SALSABILA, MOCHAMMAD ARYA (2023) Perbandingan Kinerja Algoritma Multinomial Dan Complement Naïve Bayes Untuk Analisis Sentimen Ulasan Aplikasi Traveloka Di Google Play Store. Undergraduate thesis, UNIVERSITAS PEMBANGUNAN NASIONAL "VETERAN" JAWA TIMUR.
|
Text (Cover)
19081010001-cover.pdf Download (928kB) | Preview |
|
|
Text (Bab 1)
19081010001-bab1.pdf Download (89kB) | Preview |
|
![[img]](http://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (Bab 2)
19081010001-bab2.pdf Restricted to Repository staff only until 30 May 2025. Download (268kB) |
|
|
Text (Bab 3)
19081010001-bab3.pdf Restricted to Repository staff only until 30 May 2025. Download (646kB) |
|
|
Text (Bab 4)
19081010001-bab4.pdf Restricted to Repository staff only until 30 May 2025. Download (3MB) |
|
|
Text (Bab 5)
19081010001-bab5.pdf Download (78kB) | Preview |
|
|
Text (Daftar pustaka)
19081010001-daftarpustaka.pdf Download (83kB) | Preview |
Abstract
Traveloka as the most popular online travel application in Indonesia, of course, receives various criticisms and suggestions from its users. This can be seen from the many user reviews of the Traveloka application on the Google Play Store. This study implements the Multinomial Naïve Bayes and Complement Naïve Bayes algorithms to perform sentiment analysis on 121,378 user reviews of the Traveloka application. The purpose of this study is to determine the performance of the two algorithms in classifying reviews into positive or negative sentiment categories. Multinomial Naïve Bayes works by calculating the probability of the word appearing in the document, while Complement Naïve Bayes works by calculating the probability of the word appearing in all classes except the class being observed. The test results show that the ratio of train and test data of 70%:30% gives the best performance for the two algorithms tested, and Complement Naïve Bayes outperforms Multinomial Naïve Bayes in terms of accuracy, with consecutive accuracy values of 86.46% and 84, 75%. In addition, the use of the norm True parameter in Complement Naïve Bayes gives better performance than the False norm, with an accuracy value of 88.22% and 86.46% respectively
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | T Technology > T Technology (General) | ||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Informatics | ||||||||||||
| Depositing User: | Mochammad Arya Salsabila | ||||||||||||
| Date Deposited: | 06 Jun 2023 05:05 | ||||||||||||
| Last Modified: | 06 Jun 2023 05:05 | ||||||||||||
| URI: | http://repository.upnjatim.ac.id/id/eprint/14169 |
Actions (login required)
 |
View Item |