Neural Models for Integrating Prosody in Spoken Language Understanding