Transfer Learning and Transformer Architecture for Financial Sentiment Analysis

TL;DR

This paper introduces a transfer learning-based transformer model tailored for financial sentiment analysis, aiming to improve performance with limited labeled data, especially considering pandemic-related data challenges.

Contribution

It develops a specialized pre-trained transformer model for financial sentiment analysis that leverages transfer learning and fine-tuning on small datasets, including pandemic-related data.

Findings

Effective sentiment analysis with limited labeled data

Improved model performance on pandemic-related financial data

Demonstrated adaptability of transfer learning in finance domain

Abstract

Financial sentiment analysis allows financial institutions like Banks and Insurance Companies to better manage the credit scoring of their customers in a better way. Financial domain uses specialized mechanisms which makes sentiment analysis difficult. In this paper, we propose a pre-trained language model which can help to solve this problem with fewer labelled data. We extend on the principles of Transfer learning and Transformation architecture principles and also take into consideration recent outbreak of pandemics like COVID. We apply the sentiment analysis to two different sets of data. We also take smaller training set and fine tune the same as part of the model.