Sequence-to-Sequence Data Augmentation for Dialogue Language Understanding

TL;DR

This paper introduces a sequence-to-sequence data augmentation method for dialogue language understanding that leverages semantic alternatives and diversity ranking to improve model performance with limited data.

Contribution

The paper proposes a novel data augmentation framework using sequence-to-sequence generation with diversity rank, enhancing dialogue understanding with limited training data.

Findings

Achieved 6.38 and 10.04 F-score improvements on two datasets.

Generated diverse utterances that improve language understanding.

Effective with only hundreds of training utterances.

Abstract

In this paper, we study the problem of data augmentation for language understanding in task-oriented dialogue system. In contrast to previous work which augments an utterance without considering its relation with other utterances, we propose a sequence-to-sequence generation based data augmentation framework that leverages one utterance's same semantic alternatives in the training data. A novel diversity rank is incorporated into the utterance representation to make the model produce diverse utterances and these diversely augmented utterances help to improve the language understanding module. Experimental results on the Airline Travel Information System dataset and a newly created semantic frame annotation on Stanford Multi-turn, Multidomain Dialogue Dataset show that our framework achieves significant improvements of 6.38 and 10.04 F-scores respectively when only a training set of hundreds utterances is represented. Case studies also confirm that our method generates diverse utterances.