HKUST Business School Magazine

Biz@HKUST 59 Developed by CBSA, FinSent is an interactive financial sentiment analysis web portal that automatically tracks sentiment expressed in financial reports and earnings calls of companies listed in Hong Kong and the United States. Using leading-edge Natural Language Processing (NLP) technology, this free-to-use service offers investors a fresh perspective for analyzing listed companies. It goes beyond financial information by classifying the underlying sentiments in company filings or earnings calls and presenting them in a user-friendly format. For each enquiry, FinSent returns the percentages of positive, negative and neutral sentiments, showing changes of sentiments over time. Users can also compare net sentiment percentages between companies. FinSent is backed by FinBERT, a state-of-the-art deep-learning-based NLP model developed by HKUST for analyzing financial data. FinBERT outperforms Google’s BERT model in accuracy when classifying sentiments in financial text. Investment industry practitioners from the major investment firms and financial institutions who have used the FinBERT model found it very accurate in classifying sentiments in their proprietary dataset. Project Showcase: FinSent Tracking Sentiment of Listed Companies’ Disclosure Data analytics has become a key field in the understanding of human behavior Wisers-HKUST Tourism Index FinSent financial sentiment analysis web portal The FinSent project is led by Professor Allen HUANG, Associate Dean of HKUST Business School and CBSA Associate Director (left) and Associate Professor YANG Yi of the Department of Information Systems, Business Statistics and Operations Management.