FinSentinel: Chinese Financial NLP Pipeline & Cros开源项目

FinSentinel is an end-to-end natural language processing pipeline designed to extract actionable quantitative signals from Chinese financial news. Unlike typical projects that rely on public datasets, this pipeline introduces a novel data construction method using DeepSeek-V3 (via API) for automated zero-shot annotation. It features a comprehensive benchmark comparing traditional machine learning (TF-IDF + Logistic Regression) against deep learning (Chinese FinBERT).