Comparative Analysis Of Convolutional Neural Networks, Long Short-Term Memory Networks, and Bert For Text-Based Emotion Classification
| dc.contributor.author | Visutsak, Porawat | |
| dc.contributor.author | Tongbai, Tanajak | |
| dc.contributor.author | Ongrungruaeng, Duongduen | |
| dc.contributor.author | Phongwuttisak, Nuttiruj | |
| dc.contributor.author | Wiriya, Surapong | |
| dc.contributor.author | Ryu, Keun Ho | |
| dc.date.accessioned | 2026-08-06T10:55:38Z | |
| dc.date.available | 2026-08-06T10:55:38Z | |
| dc.date.issued | 2026-06-01 | |
| dc.description.abstract | This paper presents a detailed experimental study and comparative analysis of three popular deep learning architectures (CNN, LSTM, and BERT) for emotion classification in written messages. Using a publicly available dataset of six unique emotional states (sadness, joy, love, anger, fear, and surprise), an effective ablation study was conducted to determine optimal architectural configurations, including a sequence length of 66 tokens and an embedding size of 200. To validate the results of a comparative analysis of model performance, a bootstrap technique (30 trials) and the Wilcoxon Signed-Rank test were used to eliminate potential bias. As shown by experiments, the tuned BERT architecture (with a learning rate of 2e-5) produced the most accurate and reliable result of 93.50% in classifying emotional states from texts. Moreover, with an appropriate sequence length configuration, the LSTM network (89.92%) significantly outperformed the CNN (89.65%), confirming the need to account for long-range dependencies in emotion classification. Overall, the research results show the key importance of hyperparameter tuning and the ability to handle complex information for emotion identification. | |
| dc.identifier.citation | Asean Journal of Scientific and Technological Reports, 29(6), 2026 | |
| dc.identifier.doi | 10.55164/ajstr.v29i6.262378 | |
| dc.identifier.issn | 27738752 | |
| dc.identifier.other | 2-s2.0-105040966011 | |
| dc.identifier.uri | https://dspace.kmitl.ac.th/handle/123456789/18137 | |
| dc.source | Asean Journal of Scientific and Technological Reports | |
| dc.subject | BERT | |
| dc.subject | convolutional neural networks (CNNs) | |
| dc.subject | deep learning | |
| dc.subject | Emotion classification | |
| dc.subject | long short-term memory (LSTM) | |
| dc.title | Comparative Analysis Of Convolutional Neural Networks, Long Short-Term Memory Networks, and Bert For Text-Based Emotion Classification | |
| dc.type | Article |