With the rapid growth of the OTT industry, streaming service providers and content distributors are offering consumers hundreds of thousands of video content options. To help users navigate this wealth of choice, recommendation systems are becoming increasingly important for Internet services that offer products or content. However, to provide effective recommendations, these systems need to take a multi-faceted approach, using flexible data and algorithm handling based on real-time data.

The authors Karol Chęciński and Radosław Wawrzusiak discuss the development of recommendation systems for streaming services from the perspective of both users and content platform owners in the article “Przegląd Telekomunikacyjny i Wiadomości Telekomunikacyjne”. Using a VOD service as an example, they examine the key factors that influence the quality of recommendations and the data needed to generate them. The article also addresses existing problems related to building effective recommendation systems, including theoretical and practical issues such as the implementation of recommendation algorithms in real systems.

Redge Media Recommender was developed by our R&D department to improve the streaming experience. It is a highly configurable recommendation engine capable of delivering personalized content to users in various scenarios. It combines collected knowledge about products, users and their activities and uses machine learning models tailored to the characteristics of the available data. These models take into account not only the relevance to preferences but also the variety and novelty of the suggested content to ensure long-term user satisfaction.

Redge Media Recommender was developed in collaboration with ActumLab as part of the RPO WM 2014-2020 programme.

Download the article here.

„Systemy rekomendacji – remedium w cyfrowej klęsce urodzaju” [in] „Przegląd Telekomunikacyjny i Wiadomości Telekomunikacyjne 2/2023”

Authors:

Karol Chęciński – Redge Technologies

Radosław Wawrzusiak – Redge Technologies