Recommendation Systems for Video Streaming Services: Analysis and Comparison of Algorithms in the Context of Economic Efficiency

Recommendation systems are widely used on the web to provide personalised suggestions for products and services. Streaming services are an area where recommendation engines are particularly common, as they play an important role in creating personalised experiences. In addition, improving recommendation algorithms increases the number of new users and the sale of monthly subscriptions to the service, which are important indicators of the economics of online platforms. This study analyses and compares existing recommendation algorithms in terms of their cost-effectiveness for streaming platforms. Three collaborative filtering algorithms were selected for analysis: Singular Value Decomposition (SVD), Non-Negative Matrix Factorization (NMF), and k-nearest neighbors algorithm (KNN). Different evaluation methods of recommendation systems, such as RMSE, MSE, MAE, Precision-k and Recall-k are used to determine the most efficient model. The economic efficiency of the selected recommendation systems is evaluated by analysing their ratings and their impact on the economic performance of the Netflix online cinema.