This report presents the development process and operational details of a stock market dashboard designed to combine quantitative financial data with qualitative sentiment analysis, thereby providing investors with a holistic view of market conditions. The dashboard serves as a tool for investors to make informed decisions by examining both numerical market data and the prevailing public sentiment towards the top stocks in the market.
After the Chinese stock market's steep decline in Jan, investors face a dilemma: has market sentiment bottomed out, signaling a time to buy the dip? If news headlines are overwhelmingly positive yet stock forum sentiments are bleak, it could signal an undervalued market ripe for investment. However, sifting through forums for individual comments is a time-consuming and inefficient process, and one might miss out on crucial sentiment indicators. Enter our dashboard, a fusion of real-time financial data and sentiment analysis, designed for the modern stock market. It navigates through numerical trends and investor psychology, offering a holistic view that guides investment strategies. This tool is a game-changer for investors aiming to make informed decisions in a fluctuating market!
The purpose of the dashboard is to provide a nuanced view of the stock market by integrating quantitative financial data with sentiment analysis.
The development of this dashboard used the following tools and data sources:
-
Quantitative Financial Data: Sourced from the Wind API, which offers extensive financial data for market analysis.
-
Sentiment Data: Gained through Python-based web scraping
- Investor Sentiment: Gained investor comments from the Guba and Snowball forums, focusing on the CSI 300 index stocks.
- Fund Sentiment: Captures broader market sentiment commentary from the top 100 funds in China reflecting macro investor sentiment.
- News Sentiment: Collected from Guba financial news publications to incorporate objective sentiment influences on the market.
-
Sentiment Analysis Model: Utilizes Baidu's Large Language Model (LLM) API for sentiment judgement, chosen for its accessibility and cost-effectiveness despite slower processing times.
-
Data Processing and ETL: Carried out with Python to ensure efficient data extraction, transformation, and loading into the system.
-
Database and Cloud Integration: Managed via Alibaba-cloud-based SQL databases to facilitate data handling and collaboration.
-
Visualization and Dashboard Interface: Constructed with Power BI for its dynamic data visualization capabilities, enhancing the interactive experience for users.