Article Per Year (5 Year)
p-Index From 2020 - 2025
0.23
P-Index
This Author published in this journals
All Journal Teknika
Fauzi, Ariq Ammar
Unknown Affiliation
Computer Science & IT Electrical & Electronics Engineering Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Sentiment Analysis On Tripadvisor Travel Agent Using Random Forest, Support Vector Machines, and Naïve Bayes Methods Fauzi, Ariq Ammar; Vitianingsih, Anik Vega; Kacung, Slamet; Maukar, Anastasia Lidya; Wati, Seftin Fiti Ana
Teknika Vol. 14 No. 1 (2025): March 2025
Publisher : Center for Research and Community Service, Institut Informatika Indonesia (IKADO) Surabaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.34148/teknika.v14i1.1198

Abstract

TripAdvisor faces problems in improving the quality of service on its application, namely the presence of unexpected or non-functional features, which can affect the user experience and reduce trust in the application.  This research aims to develop an application capable of performing sentiment analysis on TripAdvisor application user reviews on the Google Play Store with negative, positive, and neutral classifications using the Random Forest (RF), Support Vector Machine (SVM), and Naïve Bayes (NB). The RF method was chosen in this study because of its ability to handle large and complex data very accurately, while SVM is able to classify data on a large scale and is resistant to overfitting, while NB is able to classify text with clear probabilities. The Lexicon-based method as data labelling. The results of sentiment analysis from 1500 reviews with web scrapping show the classification of positive, negative, and neutral sentiments of 48, 726, and 646 data, respectively. Model performance in RF, SVM, and NB testing gets an accuracy value of 94%, 93.6%, and 77.8%, respectively. The RF model produces the best accuracy compared to other methods. The RF model produces the best accuracy compared to other methods. The results of sentiment analysis from 1500 user reviews allow developers to identify features that are often criticized or do not function properly in their application services.
Co-Authors Anastasia Lidya Maukar Anik Vega Vitianingsih Slamet Kacung, Slamet Wati, Seftin Fiti Ana