Yandex Unveils World's Largest Event Dataset for Recommendation Systems
2025-05-30 / Read about 0 minute
Author:小编   

Yandex has proudly launched Yambda, the most extensive open dataset for recommendation systems globally. Comprising nearly 5 billion anonymized user interaction records from its music streaming service, Yandex Music, Yambda stands as a versatile benchmark for testing cutting-edge methods and algorithms in recommendation systems. Its applicability spans diverse domains, including e-commerce, social networks, and short video platforms. The dataset, available in the Apache Parquet format, encapsulates both implicit and explicit feedback types, along with timestamps to facilitate temporal analysis. The introduction of Yambda is poised to propel advancements in recommendation system technology and expedite the pace of innovation.