Business Intelligence and Data Warehouse Approach for Analyzing Top Twitch Streamers

Abstract

In the rapidly evolving digital economy, livestreaming platforms have become massive repositories of complex, multidimensional data, presenting significant challenges for effective performance analysis. This paper proposes a new data warehouse framework to analyze the performance of streamers on the Twitch platform. Instead of using traditional flat tables that are often difficult to use for complex data, we developed a hybrid schema and an ETL process to manage many-to-many relationships between streamers and content. According to the findings, viewer distribution is very unequal, as a small group of top streamers receives most of the audience's attention. Furthermore, the study shows that geographic factors and major e-sports events have a significant impact on viewership numbers. Overall, this system provides useful information for platform operators and creators to improve their strategies in the digital economy.