|
Classifying Sentiments About Deposit Product from Social Media Texts Using Machine Learning Techniques |
|---|---|
| รหัสดีโอไอ | |
| Creator | Ploypatcha Chayanonmasakul |
| Title | Classifying Sentiments About Deposit Product from Social Media Texts Using Machine Learning Techniques |
| Contributor | Winai Nadee |
| Publisher | MSMIS Thammasat University |
| Publication Year | 2568 |
| Journal Title | Journal of information systems in Business |
| Journal Vol. | 11 |
| Journal No. | 2 |
| Page no. | 32 |
| Keyword | Emotion Classification, Comments, Social Media, Multi-label Classification, BERT Model, Deposit Products |
| URL Website | http://www.jisb.tbs.tu.ac.th |
| ISSN | 3088-1692 |
| Abstract | This research aims to develop a model for classifying emotions expressed in user comments related to deposit products. A total of 7,245 comments were collected from the official Facebook pages of commercial banks and Specialized Financial Institutions (SFIs) in Thailand. The comments, originally written in Thai, were translated into English and used to fine-tune the BERT-base-uncased model for multi-label emotion classification across ten categories: Joy, Anger, Sadness, Fear, Trust, Disgust, Surprise, Anticipation, Positive, and Negative After training the model for five epochs, the results demonstrated strong performance, with a Precision of 0.9194, Recall of 0.8526, F1-score of 0.8847, ROC AUC of 0.9614, and Accuracy of 0.6796. These findings suggest that deep learning techniques can be effectively applied to emotional analysis of customer feedback. The model can be integrated into real-world applications such as sentiment dashboards, automated alert systems for negative feedback, or marketing campaigns that respond appropriately to users' emotional tones. |
