Twitter announced a new program called Twitter Data Grants that allows research groups to request access to raw extracts of Twitter data. This basically means that researchers can analyze and process billions (possibly trillions) of tweets. In their blog post they mention possibly analyzing the data to find patterns in influenza. This sounds like a great idea, but I do wonder how valuable the data contained in tweets are. Looking at my own stream of tweets, most are just links to news articles. Then there are a handful of the “look at me” tweets that are simply people posting their personal thoughts like “it’s cold outside” or “I feel fat” or “here’s what I ate for lunch”. Maybe someone else can come up with a way to process those into something useful, but I personally just see them as people wanting to share all their thoughts with the world (I do wonder if people post these thoughts out of vanity or because they people to interact with them or maybe both). And then there’s all the bots and automatic tweets that get published from various sources. I can’t imagine what value automated tweets would provide. Of course, you could probably look for very specific things (like the flu thing) and try to analyze the timeline of tweets related to specific topic and possibly the geo of the tweets to figure out a pattern. I wonder if Twitter will share some of the proposals because I would be curious to find out what people are going to try to do with just a raw dump of data.
Well, I’ll end this post with an idea I did have about using Twitter data to identify when users do things like get a hair-cut, an oil change, go grocery shopping, get gas, etc. – things that people do at specific time intervals (like I get a hair-cut every 4 weeks). I thought if you could mine that data and then maybe a week before they’re supposed to do something, you could try to sell them on a Groupon/LivingSocial deal for a hair-cut, oil change, etc. – I figure it could be a useful marketing tool, just a lot of work and I’m not sure if you message people on Twitter like that, if they would respond. Thoughts?