KaydedilenlerDerginizi ekleyin

Dergiler / NATURENGS / 2025 / Volume: 6 Issue: 2

Sustainability-Aware Evaluating of Turkish Sentiment Analysis: Performance, Efficiency of BERTurk, DistilBERTurk, and ConvBERTurk on Humir

Bilge Koç(MALATYA TURGUT ÖZAL ÜNİVERSİTESİ, MÜHENDİSLİK VE DOĞA BİLİMLERİ FAKÜLTESİ, BİLGİSAYAR MÜHENDİSLİĞİ BÖLÜMÜ)0009-0005-4278-6640
Dilber Çetintaş(MALATYA TURGUT ÖZAL ÜNİVERSİTESİ)0000-0003-0710-2280
Sayfa
—
Yayın Tarihi
2025-12-30
🏛️Bu makalenin bibliyografik verileri ve tam metin bağlantısı DergiPark üzerinden sağlanmıştır.
DergiPark Kaydı ↗
✨

Yapay Zeka Hızlı Bakış & Çıkarımlar

AI Insight
⚡ TL;DR(1 Cümlede Hızlı Çıkarım)

Overall, the findings indicate that Turkish morphological richness can be effectively modeled by modern pretrained architectures, and the proposed evaluation framework supports sustainable model selection by jointly considering accuracy, resource consumption, and carbon impact.

📌 3 Maddede Bu Makale

🎯
Amaç & Hipotez:The performance gains of natural language processing models often come at the cost of high energy consumption and increased carbon footprint.
📊
Temel Bulgular & Veri:The models are fine-tuned on the Humir-Sentiment dataset and evaluated using accuracy, precision, recall, F1, and Matthews Correlation Coefficient (MCC).
💡
Sonuç & Katkı:Overall, the findings indicate that Turkish morphological richness can be effectively modeled by modern pretrained architectures, and the proposed evaluation framework supports sustainable model selection by jointly considering accuracy, resource consumption, and carbon impact.
🔬Tür:Modelleme & Sayısal Simülasyon
📍Kapsam:Genişletilmiş Akademik Örneklem
⚙️Değişkenler:BERTurk · ConvBERTurk · DistilBERTurk

Abstract

The performance gains of natural language processing models often come at the cost of high energy consumption and increased carbon footprint. Therefore, model evaluation should not only consider predictive accuracy but also resource usage and sustainability dimensions. This study compares pretrained Turkish language models—BERTurk, DistilBERTurk, and ConvBERTurk—for sentiment analysis through a sustainability-aware lens. The models are fine-tuned on the Humir-Sentiment dataset and evaluated using accuracy, precision, recall, F1, and Matthews Correlation Coefficient (MCC). Results show that BERTurk achieves the strongest predictive scores across accuracy, precision, recall, and F1, but at the cost of higher memory usage and longer runtimes, implying a larger estimated carbon impact. ConvBERTurk provides a balanced accuracy–efficiency profile, maintaining competitive performance while moderating resource demand. DistilBERTurk delivers the smallest memory footprint and lowest latency, translating to the lowest estimated Carbon Dioxide (CO₂) and making it well suited for resource-constrained or carbon-budgeted deployments. Overall, the findings indicate that Turkish morphological richness can be effectively modeled by modern pretrained architectures, and the proposed evaluation framework supports sustainable model selection by jointly considering accuracy, resource consumption, and carbon impact.

🔖

Bu Makaleye Atıf Yap

Koç, B., Çetintaş, D. (2025). Sustainability-Aware Evaluating of Turkish Sentiment Analysis: Performance, Efficiency of BERTurk, DistilBERTurk, and ConvBERTurk on Humir. NATURENGS, Volume: 6 Issue: 2. https://doi.org/10.46572/naturengs.1776538

📄 Tam Metin PDF Okuyucu

📄

PDF görüntüleyici hazırlanıyor...

Kaynakça (24)

  1. Wang, D., & Shao, X. (2024). Research on the impact of digital transformation on the production efficiency of manufacturing enterprises: Institution-based analysis of the threshold effect. International Review of Economics & Finance, 91, 883-897.
  2. Lee, J. D. M. C. K., & Toutanova, K. (2018). Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805, 3(8), 4171-4186.
  3. Schweter, S. (2020). BERTurk - BERT models for Turkish. *Zenodo*. https://doi.org/10.5281/zenodo.3770924
  4. Reimers, N. Sentence Embeddings using Siamese BERT-Networks. Published online August 27, 2019. doi: 10.48550. arXiv.
  5. Kemaloğlu, N., Küçüksille, E., & Özgünsür, M. (2021). Turkish sentiment analysis on social media. Sakarya University Journal of Science, 25(3), 629-638.
  6. Niu, J., Lu, W., & Penn, G. (2022, October). Does BERT rediscover a classical NLP pipeline?. In Proceedings of the 29th International Conference on Computational Linguistics (pp. 3143-3153).
  7. Devlin, J., Chang, M. W., Lee, K., & Toutanova, K. (2019, June). Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 conference of the North American chapter of the association for computational linguistics: human language technologies, volume 1 (long and short papers) (pp. 4171-4186).
  8. Kesgin, H. T., Yuce, M. K., & Amasyali, M. F. (2023). Developing and evaluating tiny to medium-sized Turkish BERT models. arXiv preprint arXiv:2307.14134.
  9. Açıkalın, U. U., Bardak, B., & Kutlu, M. (2020). BERT modeli ile türkçe duygu analizi. In 2021 28th Signal Processing and Communications Applications Conference, SIU 2020-Proceedings. Institute of Electrical and Electronics Engineers Inc.
  10. Yildirim, S. (2024). Fine-tuning transformer-based encoder for turkish language understanding tasks. arXiv preprint arXiv:2401.17396.
  11. Wicaksono, D., Rozaq, H. A. A., & Boz, N. (2025). Emotion Recognition for Low-Resource Turkish: Fine-Tuning BERTurk on TREMO and Testing on Xenophobic Political Discourse. arXiv preprint arXiv:2505.12160.
  12. Alaettin. (2024). Humir-Sentiment Datasets. https://huggingface.co/datasets/Alaettin/Humir-Sentiment-Datasets. Accessed: 12/04/2025.
  13. Choi, H., Kim, J., Joe, S., & Gwon, Y. (2021, January). Evaluation of bert and albert sentence embedding performance on downstream nlp tasks. In 2020 25th International conference on pattern recognition (ICPR) (pp. 5482-5487). IEEE.
  14. Ince, S., & Sezer, E. A. (2024, November). Retweet Prediction on Earthquake Tweets. In 2024 8th International Symposium on Multidisciplinary Studies and Innovative Technologies (ISMSIT) (pp. 1-6). IEEE.
  15. Strubell, E., Ganesh, A., & McCallum, A. (2020, April). Energy and policy considerations for modern deep learning research. In Proceedings of the AAAI conference on artificial intelligence (Vol. 34, No. 09, pp. 13693-13696).
  16. Lacoste, A., Luccioni, A., Schmidt, V., & Dandres, T. (2019). Quantifying the carbon emissions of machine learning. arXiv preprint arXiv:1910.09700.
  17. Henderson, P., Hu, J., Romoff, J., Brunskill, E., Jurafsky, D., & Pineau, J. (2020). Towards the systematic reporting of the energy and carbon footprints of machine learning. Journal of Machine Learning Research, 21(248), 1-43.
  18. International Energy Agency (IEA). (2023). CO₂ Emissions from Fuel Combustion: Overview. International Energy Agency. https://www.iea.org/reports/co2-emissions-from-fuel-combustion-overview. Accessed:15/9/2025.
  19. Çetintaş, D. (2025). Efficient monkeypox detection using hybrid lightweight CNN architectures and optimized SVM with grid search on imbalanced data. Signal, Image and Video Processing, 19(4), 1-12.
  20. AlSalman, H. (2020, March). An improved approach for sentiment analysis of arabic tweets in twitter social media. In 2020 3rd International conference on computer applications & information security (ICCAIS) (pp. 1-4). IEEE.
  21. Saad, A. I. (2020, December). Opinion mining on US Airline Twitter data using machine learning techniques. In 2020 16th international computer engineering conference (ICENCO) (pp. 59-63). IEEE.
  22. Alzyout, M., Bashabsheh, E. A., Najadat, H., & Alaiad, A. (2021, May). Sentiment analysis of arabic tweets about violence against women using machine learning. In 2021 12th International Conference on Information and Communication Systems (ICICS) (pp. 171-176). IEEE.
  23. Gupta, S., Goel, M., & Rathee, N. (2021, November). Machine learning approach to classify toxic comments on social media platforms. In Proceedings of International Conference on Data Science and Applications: ICDSA 2021, Volume 2 (pp. 439-449). Singapore: Springer Singapore.
  24. Hemakala, T., & Santhoshkumar, S. (2018). Advanced classification method of twitter data using sentiment analysis for airline service. International Journal of Computer Sciences and Engineering, 6(7), 331-335.