Why Culture and Philosophy Matter for the Studies of Algorithmic Recommender Systems. An Example from the Time Prior to the Outbreak of the Russian–Ukrainian War in 2022

Abstract

Abstract This study, conducted through 2020–2022, was an attempt to model a situation in which Russian-speaking users wanted to use YouTube to inform themselves about Crimea’s annexation, and who would then progress by watching videos relevant to the question, picking only relevant recommendations from the list of recommendations made by YouTube throughout the three stages of the experiment. The aim of the data collection was to explore filter bubbles, maintained by YouTube’s recommender algorithm, concerning this polarising issue. As a result, a network was detected which showed recommendations relevant to the topic and which was dominated by videos with positive evaluations of the Crimean annexation. A characteristic performance feature of this network is strong mutual relations, as they frequently related back to each other. As there are dependencies between the research design and the outcome, a conceptual discussion of methods in the field was included as well as the claim to broaden the perspective and to integrate the concepts of philosophy, sociology, and culture studies. In this regard, I follow Andreas Reckwitz’s understanding of the Web as ‘culture machine’ and adapt his notion of personalisation as singularisation as well as Elena Esposito's understanding of artificial communication between humans and algorithms.

Keywords

• YouTube
• Recommender algorithm
• Culture studies
• Singularisation
• Artificial communication