A couple of weeks ago at UCLA, a team of three Digital Humanists (Yoh Kawano, Todd Presner, and me) put together a live, interactive map of all the tweets in Cairo. It's viewable at http://egypt.hypercities.com/.
The project works by using Twitter's search API, requesting tweets within a few miles of the center of Cairo or which contain hashtags like #Egypt or #Jan25. While they are shown live on the map, another program downloads these from the internet and puts them into a database. The idea came to all of us around the same time: we thought we could get the project done in a day. It ended up taking about three days plus time for additional refinements that are still being made. In this blog post, I wanted to talk a little bit about our goals with the project.
Our purpose was twofold: to create an experience of the protests more immediate to the rest of the world than what we see on TV, and to create an archive to preserve the tweets for future work. The power of social media in the Egyptian protests has already been well-commented upon and debated. Our project doesn't attempt to engage with the debate over the effectiveness of social media in accomplishing the revolution. HyperCities Egypt was more about creating a focal point for making Egyptian Tweeters' messages accessible to the rest of the world. Newscasts provide one portal into the events of the day, and the Twitter feeds provided a different one that's much more focused on individual experience. While it's true that Twitter makes the same content available, locating the tweets on the map and animating their display makes the experience more immersive and compelling than Twitter's feeds. Locating the tweets also differentiated those coming from Cairo from those from other cities in Egypt, and the few from other areas.
Our project also compensates for the ephemerality of Twitter by archiving all the tweets we receive. Tweets disappear quickly, remaining available only through the Search API for a few days after they were posted. This way, we hope that it will be easier for anyone to go back and see what people were thinking during important moments—for example, that almost 25,000 tweets were posted in the six hours after Mubarak's evening resignation speech, which was more than was posted the entire rest of the day. Or even that some tweets were sent warning people about the threats of pro-government forces arriving.
This project examplified digital humanities for two reasons: it deals with both qualitative and quantitative approaches, and it appeals to audiences inside and outside the academy. First, as an interactive, qualitative visualization, HyperCities Egypt foregrounds personal experience rather than statistics by displaying individual tweets. It gives us both the power to focus on the individual experience and also to explore a staggering amount of information: as of my writing, we have about 300,000 tweets (including retweets). Some distant reading of these tweets would no doubt be interesting, but for the moment, our goal has beeen to focus on the display of the tweets rather than analysis.
In fact, I hope we'll have the time to do more analytical work in the future: an interesting research question the archive raises for me is the apparent absence of pro-Mubarak tweets. I have not seen a single pro-Mubarak tweet in English or Arabic. Though the revolution was popular, we saw footage of conflicts between supporters and opponents, and I can think of many conclusions we could draw about the apparent lack of social media among Mubarak supporters if the pattern I've observed is correct: were they all sponsored by the government, so they did not need Twitter to organize or broadcast their message?
Second, this is a project is on the one hand scholarly (in that we're creating and curating an archive) and yet also experiential. It's a tool that people who don't have in-depth knowledge of the situation in Cairo find interesting because it gives a certain perspective onto current events, and yet I hope that it will provide a useful archive for historians exploring the period, or, assuming the Google Maps APIs hold up, a useful tool for exploring the period. I think it offers an interesting example of one way we've tried to make scholarship accessible to the general public through digital media. I hope we'll be able to do more with our data and open it for more analysis.
Anyway, if anyone has any questions about the project, please post them in the comments—I'm happy to answer anything I can.