A Novel DeBERTa-based Model for Financial Question Answering Task

TL;DR

This paper introduces a DeBERTa-based model optimized with multi-model fusion for financial question answering, achieving high accuracy on the FinQA dataset and emphasizing interpretability in financial NLP tasks.

Contribution

The study presents a novel application of DeBERTa with multi-model fusion techniques for financial QA, improving accuracy and interpretability in financial reasoning tasks.

Findings

Achieved 68.99% execution accuracy on FinQA

Attained 64.53% program accuracy

Ranked 4th in the 2022 FinQA Challenge

Abstract

As a rising star in the field of natural language processing, question answering systems (Q&A Systems) are widely used in all walks of life. Compared with other scenarios, the applicationin financial scenario has strong requirements in the traceability and interpretability of the Q&A systems. In addition, since the demand for artificial intelligence technology has gradually shifted from the initial computational intelligence to cognitive intelligence, this research mainly focuses on the financial numerical reasoning dataset - FinQA. In the shared task, the objective is to generate the reasoning program and the final answer according to the given financial report containing text and tables. We use the method based on DeBERTa pre-trained language model, with additional optimization methods including multi-model fusion, training set combination on this basis. We finally obtain an execution accuracy of 68.99 and a program accuracy of 64.53, ranking No. 4 in the 2022 FinQA Challenge.