Coin Canary: Cryptocurrency media sentiment analysis

Coin Canary identifies cryptocurrency-related articles, categorizes them and then analyzes their sentiment i.e. how positive/negative/neutral is the information? By doing this for thousands of articles, we can generate a compelling picture of the media's sentiment across time, both for an individual crypto as well as for the market as a whole.

Hey PH crypto-enthusiasts! Like many people, I got deep into cryptocurrencies near the beginning of 2017. One thing that I learned very quickly was that the market was emotional: it responded quickly to news, both positive and negative. The correlation between price and sentiment seemed obvious, so I decided to explore the data. I discovered that this was rather hard to do. The predictive effect of sentiment analysis is tenuous, but correlations between price and sentiment do occur. While I was initially planning to make this a paid product, I've decided instead to open it up for anyone to view. Over time, perhaps, when the markets become less volatile, the predictive effect of the data may begin to emerge. Until then, you can use this as another data point in your investment strategy. Best of luck!

@greenwebfound Hey René! I’m glad you like it. I’d be interested to see what you guys are building. You should connect with me on Twitter so we can chat: @maplethorpej

Looks awesome, nice work Jerad! Interested to learn on a super high level you're calculating media sentiment.

@harrytucker Thanks, Harry! This is something I'd like to add to the site because it's certainly important for understanding the data. Once I get it written up, I'll post it here too.

@greenwebfound Hey René! I’m glad you like it. I’d be interested to see what you guys are building. You should connect with me on Twitter so we can chat: @maplethorpej

Looks awesome, nice work Jerad! Interested to learn on a super high level you're calculating media sentiment.

@harrytucker Thanks, Harry! This is something I'd like to add to the site because it's certainly important for understanding the data. Once I get it written up, I'll post it here too.