In this paper, we present a dataset based on publicly avail-able information. It contains listening histories of Spotify users, who posted what they are listening at the moment on the micro blogging platform Twitter. The dataset was derived using the Twitter Streaming API and is updated regularly. To show an application of this dataset, we imple-ment and evaluate a pure collaborative filtering based rec-ommender system. The performance of this system can be seen as a baseline approach for evaluating further, more so-phisticated recommendation approaches. These approaches will be implemented and benchmarked against our baseline approach in future works.