ComparingNetflix Content-Based Approaches on Movies and TV Shows

ComparingNetflix Content-Based Approaches on Movies and TV Shows

Netflix, Data Analysis, EDA, K-means

This study investigates global content production trends on the Netflix platform using a combination of Exploratory Data Analysis (EDA), K-means clustering, and data visualization techniques. The primary objective is to identify how different countries contribute to Netflix's content library and how content characteristicssuch as genre, duration, release year, and age ratingrelate to user preferences. The dataset, sourced from Kaggle and comprising over 8,800 titles, was cleaned and processed using Python in the Google Colab environment. EDA was applied to uncover statistical patterns and anomalies, while K-means clustering grouped countries based on content production features. The Elbow Method and Silhouette Score were used to determine the optimal number of clusters, revealing three distinct content clusters. Results show that the United States, India, and the United Kingdom dominate Netflix’s content production, with a strong prevalence of movies over TV shows and a notable emphasis on genres like international films and drama. Furthermore, content targeted at teen and young adult audiences exhibits higher distribution frequencies. This research demonstrates that combining EDA with unsupervised machine learning can yield actionable insights into content trends, production strategies, and user engagement. The findings offer valuable implications for content recommendation systems, localized content strategies, and global marketing approaches. By visualizing and clustering content metadata, Netflix and similar streaming platforms can enhance personalization and adapt their catalogs to suit regional and demographic preferences. This study also sets the foundation for hybrid models that incorporate user behavior with content features to improve recommendation accuracy and viewer satisfaction

Patandya Wisnu Suraya1

https://ijicom.respati.ac.id/index.php/ijicom/issue/view/14

International Journal of Informatics and Computation (IJICOM)

2025

Fajar bagus W

PDF

English

Text