Automate Stocks and Crypto Research with Python and Deep Learning

Built a Python Pipeline that Scrapes, Summarizes and Calculate Sentiment Score from Stock and Crypto News Articles from web.

1. The Scraping of the News Articles is done using BeautifulSoup-Python.
2. The deep learning model is developed using Hugging Face Transformer- Pegasus Financial Model.
3. The Sentiment Analysis is then done on the Summaries, and a sentiment score is generated via Transformers Pipeline classified as POSITIVE OR NEGATIVE.
4. The Final Output Csv is present in the repository.

The Scraping of News Articles is done from Yahoo Finance.

How to run this pipeline?

1. Clone or download this repository to your local machine.
2. Install the required libraries mentioned in the project.
3. Run the app.py command on your command prompt.
4. Pass the Code of the Stock/Crypto you want the analysis about in the monitored_tickers , for example: TSLA for Tesla, INFY for Infosys, etc.
5. Run the command, and the pipeline will automatically generate a CSV file containing all the details.

Usage