Why should we care about Twitter? Well, previous work by Singer et al shows that developers use Twitter in a professional setting:
In this study, David identified hot topics by finding events, grouping them and identifying the hottest categories. The ultimate goal is to create a domain-specific stream.
To get a set of users, David seeded an algorithm with a list of 100 famous twitterers, like Jeff Atwood and then expanded the set by looking at who they follow. This list is a public Their tweets were mined for 6 months, keyword based. The top-200 events that were outputted by an algorithm (I missed which one) and then manually sorted into categories, with the following results:
Look at the high value on Java Security updates and Jokes 🙂
There is a preprint available.