Thursday, November 27, 2008

Tweets in the Time of Crisis

Its been over 30 hours now since the Mumbai terrorist strikes began. I first learned about the attack through twitter, via an update on TwitterFox. Since then I have been glued to my computer to get information about the Maximum City - or #mumbai as the topic is referred to in twitter-speak.

The real-time results on twitter provided valuable links from citizen journalists on the ground - articles, videos, flickr photosets, contact information, opinions, wishes, prayers. Pretty soon these pictures and soundbites were lapped up by the real media and the #mumbai topic turned into a lovefest of social media and twitter [sic] Love in the Time of Cholera.

The #mumbai topic also turned into a deluge of retweets, rumors and rants. An information/misinformation overload resulted. Elsewhere, Google News ended up with 6,146 news articles related to the terrorist attacks. Social bookmarking sites like Digg and reddit eventually showed the news (after a 3 hour delay on Digg), albeit rendering the breaking news to the backburner, overshadowing it by the rick rolling at Macy's Thanksgiving parade.

Traditional television outlets began their riveting 24 hour coverage. News sources like IBNLive and (gasp) FOXNews were kind enough to broadcast live feeds on the internet. Indian channels had multiple feeds on replay mode till I switched the television off and went back to twitter.



Twitter could really use some improvements to broadcast relevant information in times like this.

Here are some suggestions.
  1. Aggregate statistics about links from tweets and retweets so people know the most popular links in a topic - #mumbai in this case.
  2. Display the actual URL instead of the tinyurl so people can identify the news source
  3. A Digg style voting so the wisdom of the crowds can be incorporated to identify the best tweets/links.
  4. Differentiate tweets from official news sites like bbc, cnn, abc from those of regular tweeters
  5. Identify local citizen journalists accurately using location detection methods such as IP-based lookups
Best!

No comments: