KaydedilenlerDerginizi ekleyin

Dergiler / Bilişim Teknolojileri Dergisi / 2026 / Cilt: 19 Sayı: 3

Bilimsel Metinlerde Anlamsal Özetleme: Derin Öğrenme Yaklaşımlarının Performans Analizi

Semantic Summarization of Scientific Texts: Performance Analysis of Deep Learning Approaches

Senanur Özcan(Kapadokya Üniversitesi, Bilgisayar ve Bilişim Teknolojileri Fakültesi, Yazılım Geliştirme Bölümü)0009-0009-5524-1630
Ahmet Özcan(Kapadokya Üniversitesi, Bilgisayar ve Bilişim Teknolojileri Fakültesi, Yapay Zeka ve Makine Öğrenmesi)0000-0002-1098-7078
Sayfa
315–329
Yayın Tarihi
2026-07-31
🏛️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)

Sonuçlar, BioMistral 7B modelinin ROUGE 1 değerinde 47.74, ROUGE 2 değerinde 21.08 ve ROUGE L değerinde 27.99 skorları ile en yüksek genel performansı elde ettiğini göstermektedir.

📌 3 Maddede Bu Makale

🎯
Amaç & Hipotez:Bu çalışmada biyomedikal soyutlayıcı metin özetleme görevinde alan odaklı ve genel amaçlı Transformer tabanlı modeller karşılaştırılmıştır.
📊
Temel Bulgular & Veri:BART Large CNN, PEGASUS Large, FLAN T5 XL ve BioMistral 7B modelleri, 133120 PubMed makalesinden oluşan hazır bir veri seti üzerinde eğitilmiş ve test edilmiştir.
💡
Sonuç & Katkı:Sonuçlar, BioMistral 7B modelinin ROUGE 1 değerinde 47.74, ROUGE 2 değerinde 21.08 ve ROUGE L değerinde 27.99 skorları ile en yüksek genel performansı elde ettiğini göstermektedir.
🔬Tür:Modelleme & Sayısal Simülasyon
📍Kapsam:Genişletilmiş Akademik Örneklem
⚙️Değişkenler:biyomedikal metin özetleme · soyutlayıcı özetleme · transformer modelleri

Özet

Bu çalışmada biyomedikal soyutlayıcı metin özetleme görevinde alan odaklı ve genel amaçlı Transformer tabanlı modeller karşılaştırılmıştır. BART Large CNN, PEGASUS Large, FLAN T5 XL ve BioMistral 7B modelleri, 133120 PubMed makalesinden oluşan hazır bir veri seti üzerinde eğitilmiş ve test edilmiştir. Tüm modeller 1024 token girdi ve 256 token çıktı sınırı ile eğitilmiş, performansları ROUGE 1, ROUGE 2, ROUGE L ve ROUGE Lsum metrikleri kullanılarak değerlendirilmiştir. Sonuçlar, BioMistral 7B modelinin ROUGE 1 değerinde 47.74, ROUGE 2 değerinde 21.08 ve ROUGE L değerinde 27.99 skorları ile en yüksek genel performansı elde ettiğini göstermektedir. Niteliksel değerlendirmeler, BioMistral 7B modelinin biyomedikal terminolojiye daha yüksek uyum sağladığını ve klinik bilgileri daha tutarlı biçimde özetleyebildiğini ortaya koymuştur. Bu bulgular, alan uyarlamalı ön eğitimin biyomedikal soyutlayıcı özetleme görevlerinde model performansını artırdığını göstermektedir.

Abstract

This study compares domain oriented and general purpose Transformer based models for biomedical abstractive text summarization. BART Large CNN, PEGASUS Large, FLAN T5 XL and BioMistral 7B models are trained and evaluated on a prepared dataset consisting of 133120 PubMed articles. All models are trained with a maximum input length of 1024 tokens and an output length of 256 tokens, and their performance is evaluated using ROUGE 1, ROUGE 2, ROUGE L and ROUGE Lsum metrics. The results show that the BioMistral 7B model achieves the highest overall performance with ROUGE 1 score of 47.74, ROUGE 2 score of 21.08 and ROUGE L score of 27.99. Qualitative analyses indicate that BioMistral 7B provides better alignment with biomedical terminology and produces more coherent clinical summaries. These findings demonstrate that domain adaptive pretraining improves model performance in biomedical abstractive summarization tasks.

🔖

Bu Makaleye Atıf Yap

Özcan, S., Özcan, A. (2026). Bilimsel Metinlerde Anlamsal Özetleme: Derin Öğrenme Yaklaşımlarının Performans Analizi. Bilişim Teknolojileri Dergisi, Cilt: 19 Sayı: 3, 315–329. https://doi.org/10.17671/gazibtd.1863483

📄 Tam Metin PDF Okuyucu

📄

PDF görüntüleyici hazırlanıyor...

Kaynakça (45)

  1. M. Azam, S. Khalid, S. Almutairi, H. A. Khattak, A. Namoun, A. Ali, and H. S. M. Bilal, “Current trends and advances in extractive text summarization: A comprehensive review,” IEEE Access, vol. 13, pp. 28150–28166, 2025, doi: 10.1109/ACCESS.2025.3538886.
  2. M. Zhang, G. Zhou, W. Yu, N. Huang, and W. Liu, “A comprehensive survey of abstractive text summarization based on deep learning,” Computational Intelligence and Neuroscience, vol. 2022, Article ID 7132226, pp. 1–18, Aug. 2022, doi: 10.1155/2022/7132226.
  3. J. Lee, W. Yoon, S. Kim, D. Kim, S. Kim, C. H. So, and J. Kang, “BioBERT: a pre-trained biomedical language representation model for biomedical text mining,” Bioinformatics, vol. 36, no. 4, pp. 1234–1240, 2020, doi: 10.1093/bioinformatics/btz682.
  4. R. Luo, L. Sun, Y. Xia, T. Qin, S. Zhang, H. Poon, and T.-Y. Liu, “BioGPT: generative pre-trained transformer for biomedical text generation and mining,” Briefings in Bioinformatics, vol. 23, no. 6, pp. 1–11, 2022, doi: 10.1093/bib/bbac409.
  5. Y. Gu, R. Tinn, H. Cheng, M. Lucas, N. Usuyama, X. Liu, T. Naumann, J. Gao, H. Poon, “Domain-specific language model pretraining for biomedical natural language processing”, Proceedings of the 2021 Conference on Health, Inference, and Learning (CHIL), Virtual Conference, 55–65, 2021.
  6. R. Tinn, H. Cheng, Y. Gu, N. Usuyama, X. Liu, T. Naumann, J. Gao, and H. Poon, “Fine-tuning large neural language models for biomedical natural language processing,” Patterns, vol. 4, no. 4, Article 100729, Apr. 2023, doi: 10.1016/j.patter.2023.100729.
  7. P. Verma and H. Om, “MCRMR: Maximum coverage and relevancy with minimal redundancy based multi-document summarization,” Expert Systems with Applications, vol. 120, pp. 43–56, 2019, doi: 10.1016/j.eswa.2018.11.022.
  8. A. John, P. S. Premjith, and M. Wilscy, “Extractive multi-document summarization using population-based multicriteria optimization,” Expert Systems with Applications, vol. 89, pp. 1–15, 2017, doi: 10.1016/j.eswa.2017.05.075.
  9. J. N. Madhuri, R. G. Kumar, “Extractive text summarization using sentence ranking”, International Conference on Data Science and Communication (IconDSC), Bangalore, Hindistan, pp. 112–117, 01–02 Mart 2019. A. K. Yadav, Ranvijay, R. S. Yadav, A. K. Maurya, “State-of-the-art approach to extractive text summarization: a comprehensive review”, Multimedia Tools and Applications, vol. 82, no. 1, pp. 29135–29197, 2023.
  10. A. M. Rush, S. Chopra, and J. Weston, “A neural attention model for abstractive sentence summarization,” in Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp. 379–389, 2015.
  11. A. See, P. J. Liu, and C. D. Manning, “Get to the point: Summarization with pointer-generator networks,” in Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (ACL), pp. 1073–1083, 2017.
  12. A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, I. Polosukhin, “Attention is all you need”, Advances in Neural Information Processing Systems (NeurIPS), Long Beach, CA, ABD, 5998–6008, 2017.
  13. F. A. Acheampong, H. Nunoo-Mensah, W. Chen, “Transformer models for text-based emotion detection: a review of BERT-based approaches”, Artificial Intelligence Review, vol. 54, no. 8, pp. 6177–6208, 2021.
  14. W. Ansar, S. Goswami, and A. Chakrabarti, “A survey on transformers in NLP with focus on efficiency”, 2024.
  15. J. Zhang, Y. Zhao, M. Saleh, and P. J. Liu, “PEGASUS: Pre-training with extracted gap-sentences for abstractive summarization,” in Proceedings of the 37th International Conference on Machine Learning (ICML), PMLR, vol. 119, pp. 11328–11339, 2020.
  16. R. Alsultan, A. Sagheer, H. Hamdoun, L. Alshamlan, and L. Alfadhli, “PEGASUS-XL with saliency-guided scoring and long-input encoding for multi-document abstractive summarization,” Scientific Reports, vol. 15, no. 26529, 2025, doi: 10.1038/s41598-025-11062-2.
  17. E. Daraghmi, L. Atwe, and A. Jaber, “A comparative study of PEGASUS, BART, and T5 for text summarization across diverse datasets,” Future Internet, vol. 17, no. 389, pp. 1–33, Aug. 2025.
  18. C. Benzoni, M. Langhals, M. Boeker, L. Modersohn, M. E. Maros, “Deep learning and abstractive summarisation for radiological reports: an empirical study for adapting the PEGASUS models’ family with scarce data”, https://arxiv.org/abs/2509.15419, Erişim tarihi: 15.02.2025.
  19. S. A. Rafi, N. Rahman, K. N. Islam, H. Ahmad, Optimizing Abstractive Summarization With Fine-Tuned PEGASUS, Teknik Rapor, Department of Computer Science and Engineering, Brac University, Bangladeş, 2023.
  20. K. N. Lam, T. G. Doan, K. T. Pham, J. Kalita, “Abstractive Text Summarization Using the BRIO Training Paradigm”, Findings of the Association for Computational Linguistics (ACL), Toronto, Kanada, 2023, doi: 10.18653/v1/2023.findings-acl.7.
  21. M. Lewis, Y. Liu, N. Goyal, M. Ghazvininejad, A. Mohamed, O. Levy, V. Stoyanov, L. Zettlemoyer, “BART: Denoising sequence-to-sequence pre-training for natural language generation”, https://arxiv.org/abs/1910.13461, Erişim tarihi: 12.10.2025.
  22. C. Raffel, N. Shazeer, A. Roberts, K. Lee, S. Narang, M. Matena, Y. Zhou, W. Li, and P. J. Liu, “Exploring the limits of transfer learning with a unified text-to-text transformer,” Journal of Machine Learning Research, vol. 21, no. 140, pp. 1–67, 2020.
  23. H. W. Chung, L. Hou, S. Longpre, B. Zoph, A. Tay, A. Fedus, “Scaling instruction-finetuned language models”, https://arxiv.org/abs/2210.11416, Erişim tarihi: 11.10.2025.
  24. A. G. Etemad, A. I. Abidi, M. Chhabra, “Fine-tuned T5 for abstractive summarization”, International Journal of Performability Engineering, vol. 17, no. 10, pp. 900–906, 2021.
  25. M. Wang, P. Xie, Y. Du, and X. Hu, “T5-Based Model for Abstractive Summarization: A Semi-Supervised Learning Approach with Consistency Loss Functions,” Applied Sciences, vol. 13, no. 12, p. 7111, 2023, doi: 10.3390/app13127111.
  26. T. Tawmo, M. Borah, P. Dadure, P. Pakray, Comparative analysis of T5 model for abstractive text summarization on different datasets, Teknik Rapor, National Institute of Technology Silchar, Assam, Hindistan, 2022.
  27. M. Yasunaga, J. Kasai, R. Zhang, A. R. Fabbri, I. Li, D. Friedman, and D. R. Radev, “ScisummNet: A Large Annotated Corpus and Content-Impact Models for Scientific Paper Summarization with Citation Networks,” in Proceedings of the AAAI Conference on Artificial Intelligence (AAAI-19), vol. 33, pp. 7386–7393, 2019.
  28. A. Cohan, F. Dernoncourt, D. S. Kim, T. Bui, S. Kim, W. Chang, and N. Goharian, “A Discourse-Aware Attention Model for Abstractive Summarization of Long Documents,” in Proceedings of the Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT), New Orleans, LA, USA, pp. 615–626, 2018.
  29. P. Tejaswin, D. Naik, P. Liu, “How well do you know your summarization datasets?”, https://arxiv.org/abs/2106.11388, Erişim tarihi: 14.02.2025.
  30. A. Chaves, C. Kesiku, and B. García-Zapirain, “Automatic text summarization of biomedical text data: A systematic review,” Information, vol. 13, no. 393, pp. 1–21, Aug. 2022, doi: 10.3390/info13080393.
  31. H. Hayashi, W. Kryściński, B. McCann, N. Rajani, C. Xiong, “What’s New? Summarizing Contributions in Scientific Literature”, Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics (EACL), pp. 1019–1031, Mayıs 2023.
  32. T. Kerner, “Domain-specific pretraining of language models: a comparative study in the medical field”, https://arxiv.org, Erişim tarihi: Şubat 2025.
  33. D. Van Veen, C. Van Uden, L. Blankemeier, J.-B. Delbrouck, A. Aali, C. Bluethgen, A. Pareek, M. Polacin, E. P. Reis, A. Seehofnerova, N. Rohatgi, P. Hosamani, W. Collins, N. Ahuja, C. Langlotz, J. Hom, S. Gatidis, J. Pauly, and A. S. Chaudhari, “Clinical Text Summarization: Adapting Large Language Models Can Outperform Human Experts,” Oct. 2023, doi: 10.21203/rs.3.rs-3483777/v1.
  34. I. Beltagy, K. Lo and A. Cohan, “SCIBERT: A Pretrained Language Model for Scientific Text”, https://arxiv.org/abs/1903.10676, Erişim tarihi: Kasım 2025.
  35. S. A. Lee, A. Wu, and J. N. Chiang, “Clinical ModernBERT: An Efficient and Long-Context Encoder for Biomedical Text”, https://arxiv.org/abs/2504.03964, Erişim tarihi: Kasım 2025.
  36. Y. Labrak, A. Bazoge, E. Morin, P.-A. Gourraud, M. Rouvier and R. Dufour, “BioMistral: A Collection of Open-Source Pretrained Large Language Models for Medical Domains”, https://arxiv.org/abs/2402.10373, Erişim tarihi: Kasım 2025.
  37. J. L. Léon, V. B. Nogueira and P. Quaresma, “Are Small Language Models Enough for Biomedical QA Tasks? Fine-Tuning Mistral-7B Using LoRA and PubMedQA,” in Proceedings of the 17th International Conference on Machine Learning and Computing (ICMLC ’25), Lecture Notes in Networks and Systems, vol. 1475, pp. 420–432, Aug. 2025, doi: 10.1007/978-3-031-94892-3_31.
  38. F. J. Dorfner, A. Dada, F. Busch, T. Han, D. Truhn, J. Kleesiek, M. Sushil, J. Lammert, M. R. Makowski, L. C. Adams, and K. K. Bressem, “Biomedical large language models seem not to be superior to generalist models on unseen medical data”, https://arxiv.org/abs/2408.13833, Erişim tarihi: Ekim 2025.
  39. M. Ravaut, A. Sun, N. F. Chen, S. Joty, “On context utilization in summarization with large language models”, Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (ACL), pp. 2764–2781, 2024.
  40. M. Kruse, S. Hu, N. Derby, Y. Wu, S. Stonbraker, B. Yao, D. Wang, E. Goldberg and Y. Gao, “Zero-Shot Large Language Models for Long Clinical Text Summarization with Temporal Reasoning,” in Machine Learning and Knowledge Discovery in Databases. Research Track (ECML PKDD 2024), Lecture Notes in Computer Science, vol. 15120, Springer, Cham, pp. 553–569, 2024, doi: 10.1007/978-3-031-94892-3_31.
  41. Cohan, A., Dernoncourt, F., Kim, D. S., Bui, T., Kim, S., Chang, W., & Goharian, N. (2018). A Discourse-Aware Attention Model for Abstractive Summarization of Long Documents. Proceedings of NAACL-HLT 2018, 615–621.
  42. C.-Y. Lin, “ROUGE: A package for automatic evaluation of summaries”, Workshop on Text Summarization Branches Out, Barcelona, İspanya, pp. 74–81, 2004.
  43. A. Ben Abacha and D. Demner-Fushman, “On the summarization of consumer health questions,” in Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (ACL), Florence, Italy, pp. 2228–2234, 2019.
  44. V. Gupta, P. Bharti, P. Nokhiz, and H. Karnick, “SUMPUBMED: Summarization Dataset of PubMed Scientific Articles,” in Proceedings of the ACL Student Research Workshop (ACL SRW), pp. 292–303, 2021.
  45. JCO Precision Oncology, “Evaluation of circulating tumor DNA as an early prognostic biomarker in metastatic breast cancer,” JCO Precision Oncology, vol. 4, pp. 1246–1262, Nov. 2020, doi: 10.1200/PO.20.00117. PMID: 35050782.