Sequential Dialogue Context Modeling for Spoken Language Understanding (2017-05-08T00:00:00.000000Z)

TL;DR

A novel approach for modeling dialogue context in a recurrent neural network (RNN) based language understanding system that allows encoding context from the dialogue history in chronological order and results in reduced semantic frame error rates.

Abstract

Spoken Language Understanding (SLU) is a key component of goal oriented dialogue systems that would parse user utterances into semantic frame representations. Traditionally SLU does not utilize the dialogue history beyond the previous system turn and contextual ambiguities are resolved by the downstream components. In this paper, we explore novel approaches for modeling dialogue context in a recurrent neural network (RNN) based language understanding system. We propose the Sequential Dialogue Encoder Network, that allows encoding context from the dialogue history in chronological order. We compare the performance of our proposed architecture with two context models, one that uses just the previous turn context and another that encodes dialogue context in a memory network, but loses the order of utterances in the dialogue history. Experiments with a multi-domain dialogue dataset demonstrate that the proposed architecture results in reduced semantic frame error rates.

Authors

Larry Heck

4 papers

Dilek Z. Hakkani-Tür

8 papers

Gökhan Tür

3 papers

Sequential Dialogue Context Modeling for Spoken Language Understanding

TL;DR

Abstract

Authors

References31 items

Towards Zero-Shot Frame Semantic Parsing for Domain Scaling

Multi-Domain Joint Semantic Frame Parsing Using Bi-Directional RNN-LSTM

End-to-End Memory Networks with Knowledge Carryover for Multi-Turn Spoken Language Understanding

Joint Online Spoken Language Understanding and Language Modeling With Recurrent Neural Networks

Towards End-to-End Reinforcement Learning of Dialogue Agents for Information Access

Dialog state tracking, a machine reading approach using Memory Network

Data Recombination for Neural Semantic Parsing

Learning End-to-End Goal-Oriented Dialog

Multi-domain Neural Network Language Generation for Spoken Dialogue Systems

Leveraging Sentence-level Information with Encoder LSTM for Semantic Slot Filling

Reward Shaping with Recurrent Neural Networks for Speeding up On-Line Policy Learning in Spoken Dialogue Systems

Semantically Conditioned LSTM-based Natural Language Generation for Spoken Dialogue Systems

Hierarchical Neural Network Generative Models for Movie Dialogues

A Hierarchical Recurrent Encoder-Decoder for Generative Context-Aware Query Suggestion

End-To-End Memory Networks

Using Recurrent Neural Networks for Slot Filling in Spoken Language Understanding

Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling

Memory Networks

The Second Dialog State Tracking Challenge

Zero-Shot Learning and Clustering for Semantic Utterance Classification

Comparing Stochastic Approaches to Spoken Language Understanding in Multiple Languages

Spoken Language Understanding: Systems for Extracting Semantic Information from Speech

Agenda-Based User Simulation for Bootstrapping a POMDP Dialogue System

Talking to machines (statistically speaking)

Long Short-Term Memory

Attention, Intentions, and the Structure of Discourse

Focusing and Description in Natural Language Dialogues

Interactive reinforcement learning for task-oriented dialogue management

End-to-End Reinforcement Learning of Dialogue Agents for Information Access

Machine Learning for Dialog State Tracking: A Review

Spoken language understanding - an introduction to the statistical framework

Field of Study