
FinSent: An AI-powered Text Analysis for Informed Investment Decisions
FinSent is a financial analysis tool powered by FinBERT, a finance-domain specialized NLP model from HKUST Business School that outperforms GPT-4 in terms of sentiment analysis and ESG classification.
FinSent is an advanced financial sentiment analysis platform that leverages FinBERT, a state-of-the-art natural language processing (NLP) model fine-tuned for financial communications. Developed by our research team, FinSent automatically extracts and classifies sentiment from earnings calls, financial reports, and corporate disclosures of companies listed in the United States and Hong Kong.
Key Features
Accurate Sentiment Tracking – Powered by FinBERT, a BERT-based model trained on 4.9B tokens of financial text, FinSent outperforms generic NLP models in detecting nuanced sentiment in financial language.
Free & Accessible – A user-friendly web portal designed for investors, analysts, and researchers to track market sentiment trends without cost.
Real-World Applications – Supports data-driven investment decisions, risk assessment, and market sentiment analysis with high precision.
Technical Innovation
FinBERT, the engine behind FinSent, is fine-tuned on a large-scale financial corpus, enabling superior accuracy in classifying positive, neutral, and negative tones in earnings discussions and regulatory filings. The model is openly available for research and development through https://huggingface.co/yiyanghkust/finbert-tone.
Explore FinSent at: https://finsent.hkust.edu.hk/