Improving Zero-Shot Text Matching for Financial Auditing with Large Language Models

TL;DR

This paper introduces ZeroShotALI, a zero-shot text matching system for financial auditing that combines a BERT-based retrieval step with an LLM filtering step, reducing the need for fine-tuning and annotated data.

Contribution

It presents a novel two-step approach using a domain-optimized transformer and a large language model for improved zero-shot document matching in financial audits.

Findings

Significant performance improvements over existing methods

Effective zero-shot document retrieval without extensive fine-tuning

Combines BERT-based retrieval with LLM filtering for accuracy

Abstract

Auditing financial documents is a very tedious and time-consuming process. As of today, it can already be simplified by employing AI-based solutions to recommend relevant text passages from a report for each legal requirement of rigorous accounting standards. However, these methods need to be fine-tuned regularly, and they require abundant annotated data, which is often lacking in industrial environments. Hence, we present ZeroShotALI, a novel recommender system that leverages a state-of-the-art large language model (LLM) in conjunction with a domain-specifically optimized transformer-based text-matching solution. We find that a two-step approach of first retrieving a number of best matching document sections per legal requirement with a custom BERT-based model and second filtering these selections using an LLM yields significant performance improvements over existing approaches.