Sequential Dialogue Context Modeling for Spoken Language Understanding

TL;DR

This paper introduces a sequential dialogue encoder for spoken language understanding that encodes dialogue history in order, improving semantic accuracy in goal-oriented dialogue systems.

Contribution

It proposes a novel Sequential Dialogue Encoder Network that models dialogue context chronologically, outperforming existing context models in SLU tasks.

Findings

Reduced semantic frame error rates with the proposed model

Sequential encoding outperforms non-chronological context models

Improved understanding in multi-domain dialogue datasets

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.