Clickbait Detection and Classification in Youtube Titles via Rule-Based Labeling

Öz

In this study, it is aimed to automatically classify clickbait content within an unlabeled dataset consisting of video titles from the YouTube platform. The labeling process was carried out using a rule-based scoring system developed based on clickbait indicators defined in the literature. For each title, a clickbait score was calculated by considering structural and linguistic features (numeral usage, capitalization ratio, punctuation density, emotional patterns, etc.), and classification was performed based on a determined threshold value. The reliability of the proposed labeling approach was supported by agreement analysis (Cohen’s Kappa) performed with human verifiers. Logistic Regression, Support Vector Machines, K-Nearest Neighbors, Random Forest, Gradient Boosting, and XGBoost machine learning algorithms were applied to the obtained labeled data. The models were evaluated using accuracy, precision, recall, F1 score, and ROC-AUC metrics; hyperparameter optimization was performed via the GridSearchCV method. The results revealed that ensemble-based algorithms (especially XGBoost and Random Forest) demonstrated the highest performance. Descriptive statistics and t-test findings showed that clickbait titles possess structures that are shorter, contain more punctuation marks, and are emphasized with expressions in uppercase letters. In this aspect, the study presents an applicable, explainable, and scalable approach for unlabeled text data without requiring human annotation.

Anahtar Kelimeler

• Clickbait
• Machine learning
• Natural language processing
• Rule-based labeling
• Text classification
• YouTube

Kural Tabanlı Etiketleme ile Youtube Başlıklarında Clickbait Tespiti ve Sınıflandırma Analizi

Öz

Bu çalışmada, YouTube platformundaki video başlıklarından oluşan etiketsiz bir veri kümesinde clickbait (tıklanma tuzağı) içeriklerinin otomatik olarak sınıflandırılması amaçlanmıştır. Etiketleme süreci, literatürde tanımlanan clickbait göstergelerine dayalı olarak geliştirilen kural tabanlı bir puanlama sistemi ile gerçekleştirilmiştir. Her başlık için yapısal ve dilsel özellikler (rakam kullanımı, büyük harf oranı, noktalama yoğunluğu, duygusal kalıplar vb.) dikkate alınarak clickbait skoru hesaplanmış ve belirlenen eşik değeri üzerinden sınıflandırma yapılmıştır. Önerilen etiketleme yaklaşımının güvenilirliği, insan doğrulayıcılar ile gerçekleştirilen uyum analizi (Cohen’s Kappa) ile desteklenmiştir. Elde edilen etiketli veri üzerinde Lojistik Regresyon, Destek Vektör Makineleri, K-En Yakın Komşu, Rastgele Orman, Gradient Boosting ve XGBoost makine öğrenmesi algoritmaları uygulanmıştır. Modeller, doğruluk, kesinlik, duyarlılık, F1 skoru ve ROC-AUC metrikleriyle değerlendirilmiş; GridSearchCV yöntemiyle hiperparametre optimizasyonu gerçekleştirilmiştir. Sonuçlar, ensemble tabanlı algoritmaların (özellikle XGBoost ve Random Forest) en yüksek performansı gösterdiğini ortaya koymuştur. Betimleyici istatistikler ve t-testi bulguları, clickbait başlıkların daha kısa, daha fazla noktalama işareti içeren ve büyük harfli ifadelerle vurgulanan yapılara sahip olduğunu göstermiştir. Bu yönüyle çalışma, etiketsiz metin verilerinde insan etiketlemesi gerektirmeden uygulanabilir, açıklanabilir ve ölçeklenebilir bir yaklaşım sunmaktadır.

Anahtar Kelimeler

• Clickbait
• Makine öğrenmesi
• Doğal dil işleme
• Kural tabanlı etiketleme
• Metin sınıflandırma
• YouTube

Kaynakça

1. Agrawal, A. (2016). Clickbait detection using deep learning. 2nd International Conference on Next Generation Computing Technologies (NGCT), IEEE. https://doi.org/10.1109/NGCT.2016.7877450.
2. Akgün, A. C. & İnkaya, T. (2024). Turkish clickbait news detection using explainable artificial intelligence. International Journal of Data Science and Applications (JOINDATA), 7(1), 64–80.
3. Altınel Girgin, B. (2021). CLUDS: Combining labeled and unlabeled data with logistic regression for social media analysis. Journal of Engineering Sciences and Design, 9(4), 1048–1061.
4. Bansal, S. (2021). 5-Minute Crafts: Video Clickbait Titles? Kaggle. https://www.kaggle.com/datasets/shivamb/5minute-crafts-video-views-dataset.
5. Bayat, S. & Işık, G. (2023a). Assessing the efficacy of LSTM, Transformer, and RNN architectures in text summarization. All Sciences Proceedings, 5th International Conference on Applied Engineering and Natural Sciences, Konya, Turkey.
6. Bayat, S. & Işık, G. (2023b). Evaluating the effectiveness of different machine learning approaches for sentiment classification. Journal of the Institute of Science and Technology, 13(3), 1496-1510. https://doi.org/10.21597/jist.1292050
7. Chakraborty, A., Paranjape, B., Kakarla, S., & Ganguly, N. (2016). Stop clickbait: Detecting and preventing clickbaits in online news media. 2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM), 9–16. https://doi.org/10.1109/ASONAM.2016.7752207
8. Chen, Y., Conroy, N. J., & Rubin, V. L. (2015). Misleading online content: Recognizing clickbait as “false news.” Proceedings of the 2015 ACM Workshop on Multimodal Deception Detection, 15–19. https://doi.org/10.1145/2823465.2823467

9. Fakhruzzaman, M. N. & Gunawan, S. W. (2021). Web-based application for detecting Indonesian clickbait headlines using IndoBERT. Science of Computer Programming (preprint).
10. Fok, T. Y., & Ye, N. (2024). Deep learning approach for knee point detection on noisy data. arXiv:2409.15608v1.
11. Gamage, B., Labib, A., Joomun, A., Lim, C. H. & Wong, K. S. (2021). BaitRadar: A multi-model clickbait detection algorithm using deep learning. ICASSP 2021 – IEEE International Conference on Acoustics, Speech and Signal Processing.
12. Genç, S. & Surer, E. (2019). Sosyal medyada “clickbait” haberlerin tespiti için makine öğrenme algoritmalarının kullanılması. 27th Signal Processing and Communications Applications Conference (SIU), IEEE.
13. Genç, S. & Surer, E. (2021). ClickbaitTR: Dataset for clickbait detection from Turkish news sites and social media with a comparative analysis via machine learning algorithms. Journal of Information Science, 49(2), 480–499.
14. Ghosh, S., Jhalani, S. & Oshiro, C. J. (2024). Topic modeling-driven feature engineering to enhance clickbait detection in social networks. 15th International Conference on Information, Intelligence, Systems and Applications (IISA 2024).
15. Gothankar, R., Di Troia, F. & Stamp, M. (2021). Clickbait detection in YouTube videos. arXiv preprint. https://arxiv.org/abs/2107.12791.
16. Huang, H., Liao, Z., Wei, X., & Zhou, Y. (2023). Combined gaussian mixture model and pathfinder algorithm for data clustering. Entropy, 25(6), 946.
17. Mowar, P., Jain, M., Goel, R. & Vishwakarma, D. K. (2021). Clickbait in YouTube: Prevention, detection and analysis of the bait using ensemble learning. arXiv preprint. https://doi.org/10.48550/arXiv.2106.06780.
18. Munna, M. H. & Hossen, M. S. (2021). Identification of clickbait in video sharing platforms. 2021 International Conference on Automation, Control and Mechatronics for Industry 4.0 (ACMI).
19. Naeem, B., Khan, A., Beg, M. O. & Mujtaba, H. (2020). A deep learning framework for clickbait detection on social area network using natural language cues. Journal of Computational Social Science, 3(2), 231–243. https://doi.org/10.1007/s42001-019-00055-4.
20. Potthast, M., Köpsel, S., Stein, B., & Hagen, M. (2016). Clickbait detection. In European Conference on Information Retrieval (pp. 810–817). Springer. https://doi.org/10.1007/978-3-319-30671-1_68.
21. Shang, L., Zhang, D., Wang, M., Lai, S. & Wang, D. (2019). Towards reliable online clickbait video detection: A content-agnostic approach. arXiv preprint. https://arxiv.org/abs/1907.07604.
22. Singh, S., Mittal, N., singh, H., & Oliva, D. (2023). Improving the segmentation of digital images by using a modified Otsu’s between-class variance. Multimedia Tools and Applications, 82: 40701-40743.
23. Thirumala, A. & Ferracane, E. (2023). Clickbait classification and spoiling using natural language processing. arXiv preprint. https://arxiv.org/abs/2306.14907.
24. Varshney, D. & Vishwakarma, D. K. (2021). A unified approach for detection of clickbait videos on YouTube using cognitive evidences. Applied Intelligence, 51, 4214–4235. https://doi.org/10.1007/s10489-020-02066-2.
25. Vaswani, A., Shazeer, N., Parmar, N., ve diğerleri. (2017). Attention is all you need. Advances in Neural Information Processing Systems, 30. https://papers.nips.cc/paper_files/paper/2017/hash/3f5ee243547dee91fbd053c1c4a845aa-Abstract.html.
26. Yücesoy, E. (2024). Konuşmacıları kadın, erkek ve çocuk olarak sınıflandırmada veri artırmanın performansa etkisi. Iğdır Üniversitesi Fen Bilimleri Enstitüsü Dergisi, 14(3), 974-987.
27. Zuhroh, N. A. & Rakhmawati, N. A. (2020). Clickbait detection: A literature review of the methods used. Register: Jurnal Ilmiah Teknologi Sistem Informasi, 6(1), 1–10.