Leveraging Large Language Models for Building Interpretable Rule-Based Data-to-Text Systems

TL;DR

This paper presents a method using large language models to automatically create interpretable rule-based data-to-text systems in Python, achieving better quality and efficiency than direct LLM prompts and fine-tuned models.

Contribution

The paper introduces a novel approach that leverages LLMs to automatically build fully interpretable rule-based systems, improving text quality and runtime efficiency.

Findings

Outperforms direct LLM prompts in text quality

Produces fewer hallucinations than fine-tuned BART

Operates faster with lower computational resources

Abstract

We introduce a simple approach that uses a large language model (LLM) to automatically implement a fully interpretable rule-based data-to-text system in pure Python. Experimental evaluation on the WebNLG dataset showed that such a constructed system produces text of better quality (according to the BLEU and BLEURT metrics) than the same LLM prompted to directly produce outputs, and produces fewer hallucinations than a BART language model fine-tuned on the same data. Furthermore, at runtime, the approach generates text in a fraction of the processing time required by neural approaches, using only a single CPU