Girish Karhana,
Rohan,
Indar Kumar,
- Student, School of Technology, Doon University, Uttarakhand, India
- Student, Department of CSE, GNIOT, Uttar Pradesh, India
- Student, Department of CSE, Independent Researcher, , USA
Abstract
Over-the-top (OTT) platforms like Netflix offer a vast and diverse range of movies as well as television programs, resulting in a difficult task of content organization. The unsupervised machine learning technique of clustering is an efficient approach to organizing these OTT platforms, as this technique helps in the grouping of content on the basis of their similar semantics, i.e., without using any labeled data. The study explores the use of clustering algorithms to enhance content segmentation. This paper proposes a comparative study of three machine learning techniques of clustering, namely K-means, hierarchical clustering, as well as DBSCAN, on Netflix OTT content. Different text parameters, such as name, director, actors, categories, as well as description, of the content are transformed into numerical values using the Word2Vec model, followed by principal component analysis (PCA) dimensionality reduction. The task of evaluation of the results are evaluated using the Silhouette Score, Davies–Bouldin Index, Calinski–Harabasz Index, as well as within-cluster sum of squared errors (SSE) measures. Experimental outcomes signify that the proposed K-means technique of clustering is superior to other clustering techniques in measuring the efficiency of the formed clusters, along with the fact that the DBSCAN technique is incompetent in forming efficient clusters because of the sparse nature of high-dimensional text embeddings. Such results help to identify and highlight the strengths and limitations of each clustering algorithm for OTT content analysis.
Keywords: OTT platforms, Netflix content analysis, unsupervised learning, clustering algorithms, Word2Vec embeddings, principal component analysis, content segmentation
[This article belongs to International Journal of Algorithms Design and Analysis Review ]
References
- Aggarwal CC. Recommender Systems. Cham: Springer International Publishing; 2016. doi:10.1007/978-3-319-29659-3.
- MacQueen J. Some methods for classification and analysis of multivariate observations. In: Le Cam LM, Neyman J, editors. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability. Vol. 1. Berkeley (CA): University of California Press; 1967. p. 281-297.
- Lloyd S. Least squares quantization in PCM. IEEE Trans Inf Theory. 1982;28:129-137. doi:10.1109/TIT.1982.1056489.
- Ester M, Kriegel HP, Sander J, Xu X. A density-based algorithm for discovering clusters in large spatial databases with noise. In: Proceedings of the Second International Conference on Knowledge Discovery and Data Mining. 1996. p. 226-231.
- Johnson SC. Hierarchical clustering schemes. Psychometrika. 1967;32:241-254. doi:10.1007/BF02289588.
- Rousseeuw PJ. Silhouettes: A graphical aid to the interpretation and validation of cluster analysis. J Comput Appl Math. 1987;20:53-65. doi:10.1016/0377-0427(87)90125-7.
- Davies DL, Bouldin DW. A cluster separation measure. IEEE Trans Pattern Anal Mach Intell. 1979;PAMI-1:224-227. doi:10.1109/TPAMI.1979.4766909.
- Jain AK. Data clustering: 50 years beyond K-means. Pattern Recognit Lett. 2010;31:651-666. doi:10.1016/j.patrec.2009.09.011.
- Zaïane OR, Foss A, Lee CH, Wang W. On data clustering analysis: Scalability, constraints, and validation. In: Pacific-Asia Conference on Knowledge Discovery and Data Mining. Berlin: Springer; 2002. p. 28-39.
- Han J, Kamber M, Pei J. Data Mining: Concepts and Techniques. Waltham (MA): Morgan Kaufmann Publishers; 2012.
International Journal of Algorithms Design and Analysis Review
| Volume | 04 | |
| Issue | 02 | |
| Received | 24/06/2026 | |
| Accepted | 28/07/2026 | |
| Published | 15/08/2026 | |
| Publication Time | 52 Days |