Enhancing Language Models for Financial Relation Extraction with Named Entities and Part-of-Speech

TL;DR

This paper enhances financial relation extraction by augmenting pre-trained language models with Named Entity Recognition and Part-of-Speech information, leading to improved performance on financial datasets.

Contribution

It introduces a simple strategy to incorporate NER and POS into language models for FinRE, demonstrating effectiveness through experiments.

Findings

NER and POS augmentation improves FinRE accuracy

The approach outperforms baseline models on financial datasets

Code and dataset are publicly available

Abstract

The Financial Relation Extraction (FinRE) task involves identifying the entities and their relation, given a piece of financial statement/text. To solve this FinRE problem, we propose a simple but effective strategy that improves the performance of pre-trained language models by augmenting them with Named Entity Recognition (NER) and Part-Of-Speech (POS), as well as different approaches to combine these information. Experiments on a financial relations dataset show promising results and highlights the benefits of incorporating NER and POS in existing models. Our dataset and codes are available at https://github.com/kwanhui/FinRelExtract.