It’s Christmas time. You have a string of 50 individually addressable RGB LEDs, what would you do? Well, [Barney] made a decision to try something different. He’s made a Christmas tree that shows Twitter’s current beliefs about the holiday.
Wait, what? We admit, it’s a type of odd concept, but the software application behind it is pretty cool. As it turns out Stanford University’s natural Language Processing group released the source code for their sentiment analyzer. Unlike a normal sentiment analyzer which assigns points to positive words and negative points for negative words, this actually utilizes a deep learning model which develops up a representation of entire sentences based on their structure — only problem? It was created and trained to examine motion picture reviews, not Christmas tweets.
Regardless, it still does the technique (kind of), but, it’s pretty slow. [Barney] has his fastest computer running four instances of the analyzer, which pulls Christmas tweets that have been sorted by the Twitter API — it then analyzes them, assigns the sentiment, and locations them in a second queue. He’s utilizing beanstalkd for the queuing and a Raspberry Pi to manage the lights. The result is a pretty light display whose colors represent the beliefs of incoming tweets — it’s hard to state if it’s actually successful in reflecting the opinion of the tweets, but it’s a pretty amazing concept.
Stick around after the break to see the Christmas Tweet Analyzing Tree in action — state that 5 times fast!