The Stock Story Scraper (SSS) is a text mining tool that gathers a stock's relevant articles and performs extensive sentiment analysis on them. Fall 2020 Independent Study - Allegheny College.
Draft a basic report/accompanying document for the project that describes the tool, the project motivations, the work completed, future work that could be completed, the accuracy of the tool, and possible shortcomings of the tool.
Define versions for Python packages in the requirements.txt file and fix issues in the web app that came up with the introduction of a new streamlit version.
As I near the completion of the basic version of the tool, I will need to refactor the code of the project to make it more efficient, syntactically correct, easier to test, and readable for others.
With this, there are a few things I know I need to refactor now:
CML code
Create standalone interface (contains options to run CML or UI)
Once I finalize the backend, which includes calculating some more results and outputting/inputting them, I will need to implement my Streamlit interface.
This interface will include a main/welcome page which asks the user to input the stocks and websites they want to use. Once this information is downloaded, an overview of each stock will be displayed on the screen giving insights into it's overall "health". Users can then go to different pages which display the articles and their information for each stock. More info will also be displayed from results.
Investigate what areas of the code (and project as a whole) can & should be refactored. Make tickets for these areas, these improvements will be included in Version 2.
I already looked into it a bit when creating #12. Start with those code areas first & then look elsewhere. #12 will likely be closed once the other tickets are created & this research is complete.
Try using things like Neural Networks or Linear Models to calculate the stock well being prediction. Need to update it from the basic calculation it is making now.