Around Christmas I was watching TV with the family and messing around with my netbook. It’s got Linux and Ruby on it, so I can actually work on that tiny thing. I came up with a script called
lovehate.rb: it downloads matches of love and hate on search.twitter.com, then calculates the frequency for each term and presents a "winner".
For some reason this script fascinated me: love almost always won, and I really wanted to find out what trends affected this. I made the script more generic and used it to compare other terms; eventually TweetFu.com was born.
TweetFu allows you to compare to terms to see which occurs the most often on Twitter. It then reruns the comparison every hour or so, producing an RSS feed and chart — the amount it reruns tests is based on demand, to keep Twitter’s API traffic low. It’s a lot like Google Trends, but currently only allows two terms to be compared.