Building a taste vector for music is a difficult task as music preferences are hard to specify concretely. It requires a lot of data and is not a stationary problem as it changes over time. Handcrafted heuristics and latent factors in playlist data can help in building taste vectors.