It has recently become feasible to run personal digital assistants on phones and other personal devices. In this paper we describe a design for a natural language understanding system that runs on device. In comparison to a server-based assistant, this system is more private, more reliable, faster, more expressive, and more accurate. We describe what led to key choices about architecture and technologies. For example, some approaches in the dialog systems literature are difficult to maintain over time in a deployment setting. We hope that sharing learnings from our practical experiences may help inform future work in the research community.
@article{arxiv.2308.03905,
title = {Intelligent Assistant Language Understanding On Device},
author = {Cecilia Aas and Hisham Abdelsalam and Irina Belousova and Shruti Bhargava and Jianpeng Cheng and Robert Daland and Joris Driesen and Federico Flego and Tristan Guigue and Anders Johannsen and Partha Lal and Jiarui Lu and Joel Ruben Antony Moniz and Nathan Perkins and Dhivya Piraviperumal and Stephen Pulman and Diarmuid Ó Séaghdha and David Q. Sun and John Torr and Marco Del Vecchio and Jay Wacker and Jason D. Williams and Hong Yu},
journal= {arXiv preprint arXiv:2308.03905},
year = {2023}
}