Automatic Data Expansion for Customer-care Spoken Language Understanding

TL;DR

This paper presents an efficient data expansion method for customer-care spoken language understanding that improves intent classification accuracy by leveraging n-gram features and out-of-domain data, reducing error rates significantly.

Contribution

The authors introduce a novel approach to expand in-domain data using n-grams and out-of-domain samples, enhancing SLU performance without extensive data collection.

Findings

Reduces classification error rate by 30%

Outperforms semi-supervised, TF-IDF, and embedding-based methods

Effective across diverse experimental setups

Abstract

Spoken language understanding (SLU) systems are widely used in handling of customer-care calls.A traditional SLU system consists of an acoustic model (AM) and a language model (LM) that areused to decode the utterance and a natural language understanding (NLU) model that predicts theintent. While AM can be shared across different domains, LM and NLU models need to be trainedspecifically for every new task. However, preparing enough data to train these models is prohibitivelyexpensive. In this paper, we introduce an efficient method to expand the limited in-domain data. Theprocess starts with training a preliminary NLU model based on logistic regression on the in-domaindata. Since the features are based onn= 1,2-grams, we can detect the most informative n-gramsfor each intent class. Using these n-grams, we find the samples in the out-of-domain corpus that1) contain the desired n-gram and/or 2) have similar intent label. The ones which meet the firstconstraint are used to train a new LM model and the ones that meet both constraints are used to train anew NLU model. Our results on two divergent experimental setups show that the proposed approachreduces by 30% the absolute classification error rate (CER) comparing to the preliminary modelsand it significantly outperforms the traditional data expansion algorithms such as the ones based onsemi-supervised learning, TF-IDF and embedding vectors.