shubxam / nifty-500-live-sentiment-analysis Goto Github PK

explain about how sentiment analysis process takes place

ideas:

• display a navigation trail for user when he searches for a specific Ticker.
  • nifty 500 > industrials > Industrial Products > Astral
• display relevant info (sentiment details) along with the news when user searches for a ticker.

• remove XC indices categorization
• Add categorization using nsepython data

update python files especially dashboard-generation.py to work with pandas 2.0 which currently gives an error

main.yml file currently just works. not efficiently.

• make use of efficient git commands.
• learn how to use github actions properly.

• Bring contents of jupyter notebook in sync with the python script.
• Well document it with proper comments.

• could use twitter data to enhance the sentiment score
• what to do about tickers with missing news articles
• Adding more news sources?
• Using whole article text instead of just headlines for calculating sentiment
• Updating the streamlit app
  • adding drop-downs.. but for what?
• bar charts to compare sentiments of companies for the choice of interval

Update the file to automatically copy sentiment_data.csv file from main branch and paste to streamlit branch

Currently while making requests using requests library, the stocks which have special characters in their name, don't return news data.