It used to take days for financial news to spread via radio, newspapers, and word of mouth. Now, in the age of the internet, it takes seconds. Did you know news articles are _automatically_ being generated from figures and earnings call streams? In this project, you will generate investing insight by applying [sentiment analysis](https://en.wikipedia.org/wiki/Sentiment_analysis) on financial news headlines from [Finviz](https://finviz.com). Using this [natural language processing](https://en.wikipedia.org/wiki/Natural_language_processing) technique, you will understand the emotion behind the headlines and predict whether the market *feels* good or bad about a stock. The datasets used in this project are raw HTML files for the [Facebook](https://finviz.com/quote.ashx?t=FB) (FB) and [Tesla](https://finviz.com/quote.ashx?t=TSLA) (TSLA) stocks from [FINVIZ.com](https://finviz.com/), a popular website dedicated to stock information and news.
- 1Searching for gold inside HTML files
- 2What is inside those files anyway?
- 3Extra, extra! Extract the news headlines
- 4Make NLTK think like a financial journalist
- 5BREAKING NEWS: NLTK Crushes Sentiment Estimates
- 6Plot all the sentiment in subplots
- 7Weekends and duplicates
- 8Sentiment on one single trading day and stock
- 9Visualize the single day
Director Data Science at multilayer.io
I am a scientist in the private sector. I do not like the data scientist title much though. What kind of scientist does not use data? I worked in genomics, video games and now fintech.