Haya, Fadhia (2026) Analisis Sentimen Berbasis Aspek Pada Pelayanan Hotel Di Pekanbaru Menggunakan Metode Naïve Bayes dan Logistic Regression. Other thesis, Universitas islam riau.

Text skripsi_203510487_watermark.pdf - Published Version Restricted to Registered users only Download (8MB) | Request a copy

Abstract

The growth of the hospitality industry in Pekanbaru City has led to an increasing number of customer reviews on online platforms, particularly Google Maps. These reviews contain various customer opinions regarding hotel service aspects, which can be utilized as a basis for evaluating service quality. However, the large volume of data makes manual analysis inefficient, thus requiring an automated approach through aspect-based sentiment analysis. This study aims to perform aspect-based sentiment analysis on hotel reviews in Pekanbaru and to compare the performance of the Naïve Bayes and Logistic Regression methods. The data were collected using web scraping techniques and processed through text preprocessing and TF-IDF feature extraction. To address data imbalance, the Synthetic Minority Oversampling Technique (SMOTE) was applied. The classification process was conducted using Naïve Bayes and Logistic Regression algorithms with a five-fold Stratified K-Fold Cross Validation evaluation scheme. The results indicate that Logistic Regression outperforms Naïve Bayes, achieving an average accuracy of 91.16%, while Naïve Bayes attains 87.38%. Therefore, Logistic Regression is more effective for aspect-based sentiment analysis of hotel customer reviews in Pekanbaru. The findings of this study are expected to assist hotel management in improving service quality and to serve as a reference for future research.

Actions (login required)

View Item