Summary:

• Downloads news articles by searching on http://www.news.google.com with keywords of interest and specific date ranges.
• Generates CVS files of news text
• Uses pre-trained NLP models to perform sentiment analysis of the news text.

• Is a scraper script which can search google news for a given date range using keywords of interest.
• sample data generated by the script looks like this. ( for keywords 'bitcoin cryptocurrency) alt text

• Takes csv file generated by google_news_scraper.py
• Performs sentiment analysis on each cell
• Performs flair (https://pypi.org/project/flair/), textblob (https://pypi.org/project/textblob/), and VADER (https://www.nltk.org/_modules/nltk/sentiment/vader.html) NLP processing to get sentiment scores.
• Averages score of each row ( by each metic) to get overall sentiment analysis score. Goal is to get overall sentiment score of all the news publised on the given date.
• Sample data generated at this stage looks like this. alt text

Code from https://towardsdatascience.com/web-scraping-news-articles-in-python-9dd605799558 is referenced as base to write scraper code.