# Error-driven Pruning of Language Models for Virtual Assistants

arXiv: https://arxiv.org/abs/2102.07219 · HTML: https://mlsanity.com/paper/2102.07219/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 6 of its 23 references that are in the map.
- First posted on arXiv 14 Feb 2021.

## Builds on (most cited first, 6 of 6)

- [A spelling correction model for end-to-end speech recognition](https://mlsanity.com/paper/1902.07178/) (2019)
- [SNDCNN: Self-normalizing deep CNNs with scaled exponential linear units for speech recognition](https://mlsanity.com/paper/1910.01992/) (2019)
- [Scalable language model adaptation for spoken dialogue systems](https://mlsanity.com/paper/1812.04647/) (2018)
- [Jointly Learning to Align and Convert Graphemes to Phonemes with Neural Attention Models](https://mlsanity.com/paper/1610.06540/) (2016)
- [Connecting and Comparing Language Model Interpolation Techniques](https://mlsanity.com/paper/1908.09738/) (2019)
- [Predicting Entity Popularity to Improve Spoken Entity Recognition by Virtual Assistants](https://mlsanity.com/paper/2005.12816/) (2020)

## Built on by (most cited first, 8 of 8)

- [Analysis and Tuning of a Voice Assistant System for Dysfluent Speech](https://mlsanity.com/paper/2106.11759/) (2021)
- [Synthetic Query Generation using Large Language Models for Virtual Assistants](https://mlsanity.com/paper/2406.06729/) (2024)
- [Space-Efficient Representation of Entity-centric Query Language Models](https://mlsanity.com/paper/2206.14885/) (2022)
- [Modeling Spoken Information Queries for Virtual Assistants: Open Problems, Challenges and Opportunities](https://mlsanity.com/paper/2304.13149/) (2023)
- [Transformer-based Model for ASR N-Best Rescoring and Rewriting](https://mlsanity.com/paper/2406.08207/) (2024)
- [Neural Language Model Pruning for Automatic Speech Recognition](https://mlsanity.com/paper/2310.03424/) (2023)
- [Phonetically-Augmented Discriminative Rescoring for Voice Search Error Correction](https://mlsanity.com/paper/2506.06117/) (2025)
- [Image-Seeking Intent Prediction for Cross-Device Product Search](https://mlsanity.com/paper/2511.14764/) (2025)

## Abstract (arXiv)

Language models (LMs) for virtual assistants (VAs) are typically trained on large amounts of data, resulting in prohibitively large models which require excessive memory and/or cannot be used to serve user requests in real-time. Entropy pruning results in smaller models but with significant degradation of effectiveness in the tail of the user request distribution. We customize entropy pruning by allowing for a keep list of infrequent n-grams that require a more relaxed pruning threshold, and propose three methods to construct the keep list. Each method has its own advantages and disadvantages with respect to LM size, ASR accuracy and cost of constructing the keep list. Our best LM gives 8% average Word Error Rate (WER) reduction on a targeted test set, but is 3 times larger than the baseline. We also propose discriminative methods to reduce the size of the LM while retaining the majority of the WER gains achieved by the largest LM.