Member-only story
I computed the correlation between the sentiment of millions of Tweets and Bitcoin prices
Here is what I found…
When I started this research, I had no idea what to expect as the final result. It started as an academic topic for my master of science thesis, over time it became much more. It’s the reason why I’m now writing on medium, sharing my knowledge with potential future practitioners interested in the field of data science.
Introduction
The use of social media has exponentially grown over the last decade. Websites are filled with posts of every kind and contain extremely valuable information that can contribute to positive ROI data science projects. In the chaos, the goal is to isolate the type of information one needs and transform it into formatting that algorithms can understand. For a project like this, it was all possible. I bet there are lots of questions so let’s go step by step:
“okay, I understand prices and correlations are easy to calculate, what about words, even worse, tweets? How can you transform a tweet into a number”
The solution is called sentiment analysis. I thoroughly explained the concept in this article. Sentiment analysis leverages…
