Sentiment Analysis and Trend Mapping of Hotel Reviews Using LSTM and GRU
This study explores applying Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models for sentiment analysis and trend mapping of hotel reviews, specifically focusing on customer feedback from Hotel Vila Ombak in Lombok, Indonesia. The primary objective was to leverage these advanced deep...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Informatics Department, Faculty of Computer Science Bina Darma University
2024-12-01
|
| Series: | Journal of Information Systems and Informatics |
| Subjects: | |
| Online Access: | https://journal-isi.org/index.php/isi/article/view/926 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849709201052401664 |
|---|---|
| author | Yerik Afrianto Singgalen |
| author_facet | Yerik Afrianto Singgalen |
| author_sort | Yerik Afrianto Singgalen |
| collection | DOAJ |
| description | This study explores applying Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models for sentiment analysis and trend mapping of hotel reviews, specifically focusing on customer feedback from Hotel Vila Ombak in Lombok, Indonesia. The primary objective was to leverage these advanced deep learning models to capture nuanced sentiment patterns in unstructured textual data, enhancing insights into guest satisfaction. The analysis was conducted on a dataset of 326 reviews, achieving an overall model accuracy of 91% (0.91). The results showed that while the models excelled in identifying positive sentiments, with a precision of 0.94, recall of 0.98, and F1-score of 0.96, they struggled with minority classes. Both negative and neutral sentiments exhibited 0% accuracy, primarily due to the dataset’s imbalance, where positive reviews constituted 92.3% of the total entries. The macro average metrics (precision 0.31, recall 0.33, F1-score 0.32) highlighted the model's limitations in classifying sentiments less frequently despite high weighted averages driven by the dominant positive class. This research underscores the need to address data imbalance and suggests that future studies incorporate techniques like data augmentation or hybrid models to improve performance across all sentiment categories. By optimizing sentiment analysis models, hospitality businesses can gain deeper insights into customer feedback, ultimately enhancing service quality and customer satisfaction. |
| format | Article |
| id | doaj-art-52340f1b8dfe4911b4fe7d110f7e0182 |
| institution | DOAJ |
| issn | 2656-5935 2656-4882 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Informatics Department, Faculty of Computer Science Bina Darma University |
| record_format | Article |
| series | Journal of Information Systems and Informatics |
| spelling | doaj-art-52340f1b8dfe4911b4fe7d110f7e01822025-08-20T03:15:24ZengInformatics Department, Faculty of Computer Science Bina Darma UniversityJournal of Information Systems and Informatics2656-59352656-48822024-12-01642814283610.51519/journalisi.v6i4.926926Sentiment Analysis and Trend Mapping of Hotel Reviews Using LSTM and GRUYerik Afrianto Singgalen0Atma Jaya Catholic University of IndonesiaThis study explores applying Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models for sentiment analysis and trend mapping of hotel reviews, specifically focusing on customer feedback from Hotel Vila Ombak in Lombok, Indonesia. The primary objective was to leverage these advanced deep learning models to capture nuanced sentiment patterns in unstructured textual data, enhancing insights into guest satisfaction. The analysis was conducted on a dataset of 326 reviews, achieving an overall model accuracy of 91% (0.91). The results showed that while the models excelled in identifying positive sentiments, with a precision of 0.94, recall of 0.98, and F1-score of 0.96, they struggled with minority classes. Both negative and neutral sentiments exhibited 0% accuracy, primarily due to the dataset’s imbalance, where positive reviews constituted 92.3% of the total entries. The macro average metrics (precision 0.31, recall 0.33, F1-score 0.32) highlighted the model's limitations in classifying sentiments less frequently despite high weighted averages driven by the dominant positive class. This research underscores the need to address data imbalance and suggests that future studies incorporate techniques like data augmentation or hybrid models to improve performance across all sentiment categories. By optimizing sentiment analysis models, hospitality businesses can gain deeper insights into customer feedback, ultimately enhancing service quality and customer satisfaction.https://journal-isi.org/index.php/isi/article/view/926lstmgrusentiment mappingtrendhotel |
| spellingShingle | Yerik Afrianto Singgalen Sentiment Analysis and Trend Mapping of Hotel Reviews Using LSTM and GRU Journal of Information Systems and Informatics lstm gru sentiment mapping trend hotel |
| title | Sentiment Analysis and Trend Mapping of Hotel Reviews Using LSTM and GRU |
| title_full | Sentiment Analysis and Trend Mapping of Hotel Reviews Using LSTM and GRU |
| title_fullStr | Sentiment Analysis and Trend Mapping of Hotel Reviews Using LSTM and GRU |
| title_full_unstemmed | Sentiment Analysis and Trend Mapping of Hotel Reviews Using LSTM and GRU |
| title_short | Sentiment Analysis and Trend Mapping of Hotel Reviews Using LSTM and GRU |
| title_sort | sentiment analysis and trend mapping of hotel reviews using lstm and gru |
| topic | lstm gru sentiment mapping trend hotel |
| url | https://journal-isi.org/index.php/isi/article/view/926 |
| work_keys_str_mv | AT yerikafriantosinggalen sentimentanalysisandtrendmappingofhotelreviewsusinglstmandgru |