Build a platform to illustrate the predicted sentiment score on given news informaiton.
Although there are many Machine Learning/Deep Learning models on stock preditions, the financial market is not rational only based on historical stock data. One major factor that influences a lot on stock market is the sentiment anlysis among fiancial news, especially for individual investors who depend a lot on the market sentiment nalysis from daily news to make portfolio allocation. However, grassively searching for different news online is exhausting and can't be gauranteed to be enough and accurate. The platform will provide the tool to gather financial news and present sentiment anlysis on the certain stock tickers(maybe chosen by users).
Backend mainly in Python. Might use Plotly Dash to illustrate the ploting and anlysis.
Stock price and sentiment are at least correlated. But we don't which factor had a primary effect on the stock market. This project is only for personal interest and academic project.
After getting the sentiment scores on the stock market, this information can be used as additonal features to improve popular models in stock prediction such as LSTM, Deep Reinforcement Learning.
Project link: https://github.com/EricaWei053/NewsSentiment
Personal website page: https://ericawei053.github.io/ericaw.github.io/
Assignments completed:
10/20: Add linting step to CI.
11/05: Set documentation, host documentation on Readthedocs, autodocumentation.
11/12: Set up personal website link to project.