Update README.md

README.md

@@ -68,10 +68,20 @@ configs:
 
 ## Overview
 
-The Yambda-5B dataset is a large-scale open database comprising **4.79 billion user-item interactions** collected from **1 million users** and spanning **9.39 million tracks**. The dataset includes both implicit feedback, such as listening events, and explicit feedback, in the form of likes and dislikes. Additionally, it provides distinctive markers for organic versus recommendation-driven interactions, along with precomputed audio embeddings to facilitate content-aware recommendation systems.
+The Yambda-5B dataset is a large-scale open database comprising **4.79 billion user-item interactions**
+collected from **1 million users** and spanning **9.39 million tracks**. The dataset includes both
+implicit feedback, such as listening events, and explicit feedback, in the form of likes and dislikes.
+Additionally, it provides distinctive markers for organic versus recommendation-driven interactions,
+along with precomputed audio embeddings to facilitate content-aware recommendation systems.
+
+The Yambda dataset is published exclusively for scientific and research purposes.
 
 Preprint: https://arxiv.org/abs/2505.22238
 
+## License
+
+[Apache 2.0](LICENSE)
+
 ## Key Features
 
 - 🎵 4.79B user-music interactions (listens, likes, dislikes, unlikes, undislikes)