The FacT: Taming Latent Factor Models for Explainability with Factorization Trees