Multi-task and multi-view learning of user state