# Bridging the Gap Between Clean Data Training and Real-World Inference for Spoken Language Understanding

arXiv: https://arxiv.org/abs/2104.06393 · HTML: https://mlsanity.com/paper/2104.06393/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 25 references that are in the map.
- First posted on arXiv 13 Apr 2021.

## Builds on (most cited first, 16 of 16)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Towards Deep Learning Models Resistant to Adversarial Attacks](https://mlsanity.com/paper/1706.06083/) (2017)
- [Domain-Adversarial Training of Neural Networks](https://mlsanity.com/paper/1505.07818/) (2015)
- [Effective Approaches to Attention-based Neural Machine Translation](https://mlsanity.com/paper/1508.04025/) (2015)
- [Improving Neural Machine Translation Models with Monolingual Data](https://mlsanity.com/paper/1511.06709/) (2015)
- [Snips Voice Platform: an embedded Spoken Language Understanding system for private-by-design voice interfaces](https://mlsanity.com/paper/1805.10190/) (2018)
- [Bridging Theory and Algorithm for Domain Adaptation](https://mlsanity.com/paper/1904.05801/) (2019)
- [Build it Break it Fix it for Dialogue Safety: Robustness from Adversarial Human Attack](https://mlsanity.com/paper/1908.06083/) (2019)
- [Understanding and Improving Lexical Choice in Non-Autoregressive Translation](https://mlsanity.com/paper/2012.14583/) (2020)
- [SlotRefine: A Fast Non-Autoregressive Model for Joint Intent Detection and Slot Filling](https://mlsanity.com/paper/2010.02693/) (2020)
- [Noisy-Labeled NER with Confidence Estimation](https://mlsanity.com/paper/2104.04318/) (2021)
- [Self-Attention with Cross-Lingual Position Representation](https://mlsanity.com/paper/2004.13310/) (2020)
- [Context-Aware Cross-Attention for Non-Autoregressive Translation](https://mlsanity.com/paper/2011.00770/) (2020)
- [Robust Multilingual Part-of-Speech Tagging via Adversarial Training](https://mlsanity.com/paper/1711.04903/) (2017)
- [Improving Robustness of Task Oriented Dialog Systems](https://mlsanity.com/paper/1911.05153/) (2019)
- [The University of Sydney's Machine Translation System for WMT19](https://mlsanity.com/paper/1907.00494/) (2019)

## Built on by (most cited first, 9 of 9)

- [Revisit Input Perturbation Problems for LLMs: A Unified Robustness Evaluation Framework for Noisy Slot Filling Task](https://mlsanity.com/paper/2310.06504/) (2023)
- [The USYD-JD Speech Translation System for IWSLT 2021](https://mlsanity.com/paper/2107.11572/) (2021)
- [Robustness of Large Language Models to Perturbations in Text](https://mlsanity.com/paper/2407.08989/) (2024)
- [Divide, Conquer, and Combine: Mixture of Semantic-Independent Experts for Zero-Shot Dialogue State Tracking](https://mlsanity.com/paper/2306.00434/) (2023)
- [PSSAT: A Perturbed Semantic Structure Awareness Transferring Method for Perturbation-Robust Slot Filling](https://mlsanity.com/paper/2208.11508/) (2022)
- [Towards Robust and Generalizable Training: An Empirical Study of Noisy Slot Filling for Input Perturbations](https://mlsanity.com/paper/2310.03518/) (2023)
- [DemoNSF: A Multi-task Demonstration-based Generative Framework for Noisy Slot Filling Task](https://mlsanity.com/paper/2310.10169/) (2023)
- [Prompt-Learning for Cross-Lingual Relation Extraction](https://mlsanity.com/paper/2304.10354/) (2023)
- [Noise-BERT: A Unified Perturbation-Robust Framework with Noise Alignment Pre-training for Noisy Slot Filling Task](https://mlsanity.com/paper/2402.14494/) (2024)

## Abstract (arXiv)

Spoken language understanding (SLU) system usually consists of various pipeline components, where each component heavily relies on the results of its upstream ones. For example, Intent detection (ID), and slot filling (SF) require its upstream automatic speech recognition (ASR) to transform the voice into text. In this case, the upstream perturbations, e.g. ASR errors, environmental noise and careless user speaking, will propagate to the ID and SF models, thus deteriorating the system performance. Therefore, the well-performing SF and ID models are expected to be noise resistant to some extent. However, existing models are trained on clean data, which causes a textitgap between clean data training and real-world inference. To bridge the gap, we propose a method from the perspective of domain adaptation, by which both high- and low-quality samples are embedding into similar vector space. Meanwhile, we design a denoising generation model to reduce the impact of the low-quality samples. Experiments on the widely-used dataset, i.e. Snips, and large scale in-house dataset (10 million training examples) demonstrate that this method not only outperforms the baseline models on real-world (noisy) corpus but also enhances the robustness, that is, it produces high-quality results under a noisy environment. The source code will be released.