Yesterday I spent some time at the College of Information Sciences and Technology’s Graduate Symposium listening to Dr. Abdur Chowdhury, Twitter’s Chief Scientist give a keynote talk. I’ve written about Twitter for a couple of years now, but what is so interesting about Twitter these days is what is going on behind the scenes. While they are a relatively new company, they are really working hard to make sense of the river of data that flows through the 140 character text box they offer. Abdur was the co-founder of Summize where they used analytics to discover trends across the web … when Twitter bought it he moved over to start decoding massive trends brought about by Twitter’s user base. Later in the day I was then on a panel with him talking about Twitter to those assembled it was very interesting to hear his perspective on things and to see how he is leading the way in making sense out of all of those Twitter messages.
As an aside, one of the questions that seems to occupy lots of cycles is about Twitter’s financial model. I could tell during the question and answer period yesterday it is something that gets the folks at Twitter a little frustrated. I recently read an article in Wired where they make the claim that Twitter can go for quite some time without worrying about that and I am betting they are figuring it out behind closed doors on their own time.
One thing he spent a lot of time on was how they can pull content from peoples’ tweets and find real news trends. He referenced Mumbai several times as an example of where Twitter was able to bring the terrorist attacks to our attention even before the news. We saw it a few weeks ago with the flight that landed in the Hudson … pictures and reports came from the Twittersphere before the news had any clue. This isn’t new for people who have been paying attention, but it was a good push for me to revisit the capabilities of Twitter Search. It seems like everyone is discovering the power Twitter has to offer.
Being on the panel reminded me of how interesting the Twitter experiment Scott McDonald and I did in the CI 597C course we taught last year. I got to talking with Abdur about it and the research we had planned to do … talking about it all has made me even more interested in using Twitter to better understand what is going on in my classroom. When I did a google search on how far back tweets go, I came across Twistory. It is a site that uses the API to pull out the exact time and content of any given user’s Twitter account. What is really cool is that you can then use Google Calendar to subscribe to the output to visualize the data.
Seeing Twitter Conversations
Now this is where it gets really interesting to me … you can put anyone’s name into it and add their twitter stream to the calendar. This set off an ah-ah moment for me. What it means is that in a class you could easily visualize the backchannel conversation between and among students. Imagine how rich the data can be now looking at what happens in class — are students passing twitter notes, digging deeper into the conversation, exchanges resources, etc? This is the first time I’ve been able to create a tangible paper trail of the interactions happening behind the scenes. I can’t wait until Spring 2010 when Scott and I teach our Disruptive Technologies course. We met this morning at the coffee shop to start talking about how we would integrate this and we’ve decided that while we are focusing on our themes of community, identity, and design that we’ll ask students to do research into how the community is coming together and evolving by mining Twitter data in this form (or another). I can’t wait to see how it goes down and to explore other ways to use Twitter and its API.
The idea of twitter as a kind of pulse of the web is a pretty powerful one. And the Mumbai example is a very telling one it that regard.
And wow, that Google Calendar/Twitter mashup is amazing, and the implications could be awesome. I’m excited to see what you do with Twitter in your class, because I have still yet to see a compelling example of how it might be imagined as part of a course, rather than an adjunct. This method marries the two beautifully in my mind, very cool stuff.
The concept of learning from streaming feedback of a speech or being able to modify lectures based on the behind-the-scenes conversation taking place during a class seems very promising as a teaching aid. It reminds me of the sort of instant polling that is conducted during political events.
Also, what would be really cool is if you were able to extract data from and organize tweets based on similar words or phrases…perhaps even chart them on a timeline. I think this sort of mash-up would provide a fascinating glimpse into the real “pulse” of a group.
Right after I read, with a bit of disbelief, that there’s a bill in Congress proposing ISPs keep user records for the police, I read your innocent and excited post. You say:
” Imagine how rich the data can be now looking at what happens in class — are students passing twitter notes, digging deeper into the conversation, exchanges resources, etc? This is the first time I’ve been able to create a tangible paper trail of the interactions happening behind the scenes”
Not so sure it’s a good thing for anyone in authority to have the power to eliminate private discussion, and there in, frustrate dissension, civil disobedience, and all the things that are so near and dear to my heart. Where’s the student’s right to laugh behind your back at your tie?
“a site that uses the API to pull out the exact time and content of any given user’s Twitter account” is a powerful, useful, and very scary thing- scary not so much because it exists, but because everybody just thinks it’s cool.
… let me quickly add that I trust your intentions completely and think your ties are fine.
@ RD Huff That is exactly what we are considering … writing something that can look really closely at the content of the captured tweets and see what kind of sense we can make out of it. Last spring when we taught this course to a collection of graduate students I would watch twitter during class and after about week 4 there was a very powerful backchannel discussion happening. Most of the time it was to ask deeper questions, push links, or engage. The other thing that went on were contributions from outside the class based on the twitter stream … it stunned Scott and I.
And to @david stong point, I agree it may seem a bit invasive, but what we are really considering is pushing the students to do the sense making themselves and let them discover connections that are happening as the learning community (in and out) of the class emerges. Like last time, we’ll leave it up to them to contribute to the discussion, but digging in and making sense of the data will be part of the requirements. I’ll also say, they twittered plenty of snarky comments. Also, try to keep in mind I try to never wear ties!
@ Jim Next time we get together we can talk about how it all played out in the last class. We didn’t require it, we discussed it as a disruptive technology and 3/4 of the class took off with it. Really made the semester work in a way I hadn’t seen before. In a lot of ways it made all of us know each other better — in a strange way. Trust me, when it goes down I’ll be writing about it!
Pingback: Why Run a Service? « Cole Camplese: Learning and Innovation