Large Language Models for Simultaneous Named Entity Extraction and Spelling Correction

TL;DR

This paper explores using decoder-only Large Language Models to simultaneously extract Named Entities and correct spelling errors in OCR-processed Japanese receipt text, demonstrating comparable performance to BERT-based models.

Contribution

It introduces a novel approach of using generative LLMs for joint NE extraction and spelling correction, extending beyond traditional classification methods.

Findings

Best LLM performs as well as BERT models in NE extraction.

LLMs can automatically correct some OCR errors.

Fine-tuned LLMs show potential for joint NE extraction and spelling correction.

Abstract

Language Models (LMs) such as BERT, have been shown to perform well on the task of identifying Named Entities (NE) in text. A BERT LM is typically used as a classifier to classify individual tokens in the input text, or to classify spans of tokens, as belonging to one of a set of possible NE categories. In this paper, we hypothesise that decoder-only Large Language Models (LLMs) can also be used generatively to extract both the NE, as well as potentially recover the correct surface form of the NE, where any spelling errors that were present in the input text get automatically corrected. We fine-tune two BERT LMs as baselines, as well as eight open-source LLMs, on the task of producing NEs from text that was obtained by applying Optical Character Recognition (OCR) to images of Japanese shop receipts; in this work, we do not attempt to find or evaluate the location of NEs in the text. We show that the best fine-tuned LLM performs as well as, or slightly better than, the best fine-tuned BERT LM, although the differences are not significant. However, the best LLM is also shown to correct OCR errors in some cases, as initially hypothesised.