SENTIMEN ANALISIS APLIKASI TRAVELOKA MENGGUNAKAN ALGORITMA SUPPORT VECTOR MACHINE

DAMAYANTI, ADELIA and Sutedi, Sutedi (2024) SENTIMEN ANALISIS APLIKASI TRAVELOKA MENGGUNAKAN ALGORITMA SUPPORT VECTOR MACHINE. Skripsi thesis, Institut Informatika dan Bisnis Darmajaya.

[img] Text
COVER ADEL.pdf
Download (105kB)
[img] Text
HAL PERSETUJUAN 1911050073.pdf
Download (502kB)
[img] Text
HAL PENGESAHAN 1911050073.pdf
Download (488kB)
[img] Text
PERNYATAAN 1911050073.pdf
Download (162kB)
[img] Text
Intisari adelia.pdf
Download (124kB)
[img] Text
ABSTRAK ADELIA.pdf
Download (9kB)
[img] Text
PRAKATA 1911050073.pdf
Download (228kB)
[img] Text
DAFTAR ISI.pdf
Download (14kB)
[img] Text
BAB I.pdf
Download (155kB)
[img] Text
BAB II.pdf
Download (114kB)
[img] Text
BAB III.pdf
Download (310kB)
[img] Text
BAB IV.pdf
Download (1MB)
[img] Text
BAB V.pdf
Download (8kB)
[img] Text
Daftar Pustaka Adel.pdf
Download (86kB)

Abstract

Many online ticketing applications in Indonesia can be used online or downloaded from the Google Play Store for Android users. Five online ticketing applications are most widely used by Android users, namely: Agoda, Mister Aladin, Pegipegi, Tiket.com, and Traveloka. Each of these online ticketing programs has advantages and disadvantages. This research aimed to see sentiment analysis on the Traveloka Application. The downside is that much fraud occurs during transactions, so when consumers cancel a ticket order or hotel reservation, it sometimes takes a long time to get a refund. This research was analyzed using the Support Vector Machine algorithm. The tool used was the Rapidminer application. The stages carried out were data collection, data pre-processing, data classification, SVM cross-validation, and accuracy value. The sentiment classification results by using 500 data obtained that 321 or 64.2% of the data entered the positive class and the remaining 179 or 35.8% entered the negative class and produced an accuracy rate of 85.8%. This means that of the 500 reviews, there were 427.5 reviews classified exactly in the sentiment class. In comparison, 14.50 or 22.5% of reviews are not correctly classified in the sentiment class.

Item Type: Thesis (Skripsi)
Subjects: Ilmu Komputer
eSkripsi
Divisions: Skripsi/TA & PKPM/KP - Fakultas Ilmu Komputer > Sistem Informasi
Depositing User: Adelia -Damayanti
Date Deposited: 17 Sep 2024 08:34
Last Modified: 17 Sep 2024 08:34
URI: http://repo.darmajaya.ac.id/id/eprint/18462

Actions (login required)

View Item View Item