Improving Top-K Decoding for Non-Autoregressive Semantic Parsing via Intent Conditioning

TL;DR

This paper introduces a novel non-autoregressive semantic parser with intent conditioning, enhancing top-k output diversity and quality while maintaining fast inference, suitable for conversational virtual assistants.

Contribution

It proposes intent conditioning in NAR semantic parsing to improve top-k output quality and diversity, addressing beam search challenges and maintaining O(1) decoding complexity.

Findings

Top-3 EM improved by 2.4 points

Speeds up beam search inference by 6.7x on CPU

Maintains competitive top-k EM with faster inference

Abstract

Semantic parsing (SP) is a core component of modern virtual assistants like Google Assistant and Amazon Alexa. While sequence-to-sequence-based auto-regressive (AR) approaches are common for conversational semantic parsing, recent studies employ non-autoregressive (NAR) decoders and reduce inference latency while maintaining competitive parsing quality. However, a major drawback of NAR decoders is the difficulty of generating top-k (i.e., k-best) outputs with approaches such as beam search. To address this challenge, we propose a novel NAR semantic parser that introduces intent conditioning on the decoder. Inspired by the traditional intent and slot tagging parsers, we decouple the top-level intent prediction from the rest of a parse. As the top-level intent largely governs the syntax and semantics of a parse, the intent conditioning allows the model to better control beam search and improves the quality and diversity of top-k outputs. We introduce a hybrid teacher-forcing approach to avoid training and inference mismatch. We evaluate the proposed NAR on conversational SP datasets, TOP & TOPv2. Like the existing NAR models, we maintain the O(1) decoding time complexity while generating more diverse outputs and improving the top-3 exact match (EM) by 2.4 points. In comparison with AR models, our model speeds up beam search inference by 6.7 times on CPU with competitive top-k EM.