ZEROTOP: Zero-Shot Task-Oriented Semantic Parsing using Large Language Models

TL;DR

ZEROTOP leverages large language models by decomposing semantic parsing into question-answering tasks, enabling zero-shot task-oriented parsing with minimal annotated data, and addressing unanswerable question detection through fine-tuning.

Contribution

The paper introduces ZEROTOP, a novel zero-shot semantic parsing method that decomposes tasks into QA problems and fine-tunes LLMs to handle unanswerable questions, achieving notable parsing accuracy.

Findings

Correctly parses ~16% of utterances in MTOP dataset without annotated data.

Uses QA decomposition to leverage LLMs for semantic parsing.

Fine-tunes LLMs to detect unanswerable questions.

Abstract

We explore the use of large language models (LLMs) for zero-shot semantic parsing. Semantic parsing involves mapping natural language utterances to task-specific meaning representations. Language models are generally trained on the publicly available text and code and cannot be expected to directly generalize to domain-specific parsing tasks in a zero-shot setting. In this work, we propose ZEROTOP, a zero-shot task-oriented parsing method that decomposes a semantic parsing problem into a set of abstractive and extractive question-answering (QA) problems, enabling us to leverage the ability of LLMs to zero-shot answer reading comprehension questions. For each utterance, we prompt the LLM with questions corresponding to its top-level intent and a set of slots and use the LLM generations to construct the target meaning representation. We observe that current LLMs fail to detect unanswerable questions; and as a result, cannot handle questions corresponding to missing slots. To address this problem, we fine-tune a language model on public QA datasets using synthetic negative samples. Experimental results show that our QA-based decomposition paired with the fine-tuned LLM can correctly parse ~16% of utterances in the MTOP dataset without requiring any annotated data.