January 28th, 2009
Which is a Better Mobile Citizen Reporting Tool: Twitter or Ushahidi? (Updated)
Welcome to Gauravonomics Blog! Subscribe to my feed now and you'll never miss a single post!
The ubiquitous access to mobile phones, and the ease of sending SMSes, has dramatically democratized citizen reporting. In a recent post on the digital news life cycle, I argued that “from this moment onwards, every accident worth reporting, anywhere in the world, will be reported first, via SMS, by a bystander who has a mobile phone.”
In 2008, we witnessed mobile-based citizen reporting play an important role in covering the earthquake in China, the terror attack in Mumbai, and the post-election violence in Kenya.
However, in all these cases, even as bloggers and journalists praised the role of mobile based citizen crisis reporting, several observers pointed out that the Twitter SMS stream had more noise than signal and it was impossible to extract meaning out of it. In the Mumbai terror attack, for instance, many of us struggled to make sense of the high volume #Mumbai feed, even after we shifted first to geo-tagged tweets from people living close to Mumbai, and then to such tweets containing links.
If a crisis is geographically distributed, it becomes even more difficult to make sense of the SMS stream.
In the case of China earthquake, John Kennedy at Global Voices put together a location-wise list of Twitter users giving updates on the situation. A manually curated list like that is useful, but only in a limited way.
In the case of Kenya’s post-election violence, Kenyan bloggers, led by Erik Hersman, called for a Google Map mashup to track the situation, quickly created such a mashup in Ushahidi, and even improved upon it quickly.
Patrick Meier and Kate Brodock did a comparison between the coverage of the post-election violence in Kenya on blogs (reporting based on SMSes), mainstream media and Ushahidi and found that none of the sources were complete in themselves, but Ushahidi had more precise location information and wider coverage. Unfortunately, Ushahidi has been much less successful in its other implementations in Congo and Gaza.
We are still unwrapping the bubble wrap here and although there is much enthusiasm about all types of mobile citizen reporting tool, including Ushahidi, Twitter and even Brightkite, there is very little research on which one of them is better for crisis reporting.
Here’s one way to think about the problem. There are two conflicting considerations involved here: location and ubiquity.
Ushahidi co-founder Erik Hersman points to his “pothole theory of digital activism” (we only care about a pothole if it’s on our street) and argues that location is the key in mobile crisis reporting, because location makes crisis information relevant to us.
As a counter-point, I would like to point to Ethan Zuckerman’s “cute cat theory of digital activism” (most of us accidentally stumble upon activism while we upload pictures of cute cats on the internet) and argue that ubiquity is the key in mobile crisis reporting, as people tend to use the same service for finding entertainment and engaging in citizen reporting.
In the end, the answer, as always, is the always unsatisfactory “it depends”.
In most cases, normal people will accidentally become citizen reporters and use tools like Twitter to share their 140 letters on a crisis. It will be up to us then to make sense of the high volume firehose of these unstructured tweets, using filters like near:location, or mashups built on top of Twitter.
Sometimes, a small percentage of us will feel motivated to find out structured citizen reporting platforms like Ushahidi or Twitter VoteReport and send in well-structured tweets with location information.
Increasingly, many of us will use a Brightkite like system to pre-declare our locations, so that we don’t have to worry about the syntax of our SMSes.
In any case, crisis reporting will always follow the 1:9:90 rule, with a small minority using dedicated citizen reporting platforms and a large majority using multi-purpose communications platforms.
Going beyond the location versus ubiquity debate, extracting meaning from a SMS firehose, whether on Twitter or Ushahidi, ultimately boils down to the robustness of social search tools built on top of them. As I said in a recent post, the ideal social search engine, however, would allow you to sort the search results by relevance, authority or recency, as per your requirements. As of now, no search engine (including Twitority and Twithority) does this well for SMS, and social search for SMS is still very much work in progress.
Which brings me back to the question which started this entire chain of thoughts. Indian elections are as prone to violence and sabotage as elections anywhere else (especially in pockets in North and East India), and I’m sure many people will be using SMS in one form or other to share information on such incidents on election day. What will be the best way to make sense of all these SMSes during the 2009 Indian Lok Sabha elections? Is anybody else, apart from me, thinking about that question?
Cross-posted at MSFS 556: Social Media in Business, Development and Government.
