A Deep Learning Approach for Emotion Detection in Indonesian Social Media Texts
DOI:
https://doi.org/10.56480/jln.v6i1.1647
Abstract View:
4
PDF downloads:
3
Keywords:
Emotion Classification, Deep Learning, CNN, BiLSTM, Indonesian LanguageAbstract
Automated emotion analysis of Indonesian social media text is challenging due to linguistic complexity and the nuanced use of emojis. This study addresses this by designing and evaluating a serial hybrid deep learning architecture combining a Convolutional Neural Network (CNN) and a Bidirectional Long Short-Term Memory (BiLSTM) network. The model was trained on an augmented Emotion Twitter dataset from the IndoNLU benchmark. Following a comprehensive preprocessing pipeline, the model achieved a promising test accuracy of 89.24% and a Macro F1-Score of 0.90 across five emotion classes. While the model excelled at identifying distinct emotions like 'love' (F1-Score of 0.96), it faced challenges in distinguishing between 'happy' and 'sadness'. These results establish the serial hybrid CNN-BiLSTM architecture as a viable and effective baseline for Indonesian emotion classification, providing a solid foundation for future research into more advanced models.
References
Madan, A., & Kumar, D. (2024). Real-time topic-based sentiment analysis for movie tweets using hybrid approach. Knowledge and Information Systems, 66(7), 3061–3083. https://doi.org/10.1007/s10115-024-02298-x
Pane, S. F., Ramdan, J., Putrada, A. G., Fauzan, M. N., Awangga, R. M., & Alamsyah, N. (2022). A hybrid CNN-LSTM model with word emoji embedding for improving the Twitter sentiment analysis on Indonesia’s PPKM policy. Dalam 2022 International Conference on Information Technology Systems and Innovation (ICITSI). IEEE. https://doi.org/10.1109/ICITSI56531.2022.10057720
Phan, H. T., Seo, Y.-S., & Nguyen, N. T. (t.t.). Fuzzy hybrid CNN-LSTM model for sentence-level sentiment analysis. SSRN. https://doi.org/10.2139/ssrn.4994871
Riyadi, S., Andriyani, A. D., & Sulaiman, S. N. (2024). Improving hate speech detection using double-layers hybrid CNN-RNN model on imbalanced dataset. IEEE Access, 12, 159660–159668. https://doi.org/10.1109/ACCESS.2024.3487433
Shaver, P. R., Murdaya, U., & Fraley, R. C. (2001). Structure of the Indonesian emotion lexicon. Asian Journal of Social Psychology, 4(3), 201–224. https://doi.org/10.1111/1467-839X.00086
Singh, A. K., Bhushan, A., & Dwivedi, D. (2024). Analyzing sentiments on Twitter using deep learning techniques. International Journal of Modern Education and Computer Science, 16(1), 60–72. https://doi.org/10.5815/ijmecs.2024.01.05
Wilie, B., Vincent, T. N., Cahyawijaya, S., Li, X., Lim, Z. Y., Sutiono, A., ... & Purwarianti, A. (2020). IndoNLU: Benchmark and resources for evaluating Indonesian natural language understanding. Dalam Proceedings of the 1st Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 10th International Joint Conference on Natural Language Processing. Association for Computational Linguistics. https://doi.org/10.18653/v1/2020.aacl-main.85
Downloads
Published
How to Cite
License
Copyright (c) 2025 Rangga Widiasmara, I Gede Susrama Mas Diyasa, Chrystia Aji Putra
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Copyright Notice
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution-ShareAlike 4.0 International License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
