A STUDY OF FINBERT'S EFFECTIVENESS IN FINANCIAL TEXT ANALYSIS: INSIGHTS INTO SENTIMENT, ESG FACTORS, AND PREDICTIVE OUTCOMES FROM MALAYSIAN FINANCIAL INSTITUTIONS
Loading...
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
UNIVERSITI MALAYSIA SARAWAK
Abstract
Description
As the economy and financial landscape evolve, the ability to process and interpret large volumes of unstructured financial information becomes increasingly critical for both institutions and individual investors. This study aims to evaluate the effectiveness of the FinBERT, a specialised NLP model fine-tuned for financial text analysis, in analysing financial texts to extract sentiment, evaluate Environmental, Social, and Governance (ESG) factors and identify forward-looking statements (FLS) from corporate annual reports of Malaysian financial institutions. A comprehensive methodology, that includes a literature review, data collection, text preprocessing, the application of the FinBERT model to conduct financial text analysis, and fine-tuning the model, is employed. The findings of this study aim to assess the model’s performance and its potential to enhance financial decision-making and bridge the gap between financial literacy and practical application.