Stocks and Crypto Research

  • Tech Stack: Python, scikit-learn, numpy, pandas, Tensorflow, keras
  • Github URL: Project Link

This research includes the News and Tweet's Sentiment Analysis and trying the predict and forcast the stock and crypto's price.

News Analysis

  • The news is scraped directly from `Google`, and all the out unwanted URLs are striped out.
  • This is scraped out news is then summarized using `Pegasus` Model.
  • `Flair` Sentiment Analysis is used to find the sentiment of the summarized News.

    • Tweet Analysis

  • The Tweet are scraped directly from Twitter using the `snscrape` module.
  • With some preprocessing steps we clean the tweets.
  • `Flair` Library is used to find the sentiment of the Tweets.
  • `Yahoo Finanace` library is used to get the historical data of the required Stock.
  • Comparing our tweet sentiments against real stock data.