MIT Laboratory for Social Machines, a bet to avoid a mayor emergency in Twitter
The MIT Media Lab announced the creation of the Laboratory for Social Machines (LSM), whose main goal would be the creation of programs aimed to develop collaborative technologies to tackle complex social problems, as you can read here. This would be great news for the emergency management world as long as Twitter itself developes new technologies capable of detecting a big streaming growing up in Twitter. Once we have this problem solve, it would turn into the best tool for PIOs to control rumors when a disaster strikes. If we admit that Twitter is the biggest web of reporters existing ever (with all their drawbacks) and that it offers a synchronized experience, as @dkroy explained in #TATGranada2014, there is an urgent need to put an order in this huge stream of information. We’ll try to explain in this post (co written with @LuisSerranoR) why and how this new tool may help.
It is true that there are some ways to monitor conversation in social media (we are talking about Twitter basically here) when a disaster strikes. We can add a column either in Hoosuite or Tweecked with a Hashtag (ie: #Sandy, #yycflood) and this helps to monitor the conversation. I always tell my students that a tweet without a Hashtag is like going to a Big Store without a directory indicating you where the goods are. If you want to buy shoes you go to the shoe section- not to the bakery. So if you want to follow the news of a Hurricane your follow the HT with its name.
We could also follow the official source of information (usually the Government, local authority, and so on) or a list of influencers and people we trust as a reliable source of information, but still, we cannot stop rumors because we lack relevant information: where and when (the exact time) the tweet was posted being the more important. Moreover, we need to detect the rumor going viral from the very beginning.
Time is relevant to stop rumors going viral
The thing is we don’t have any app capable of detecting a huge stream of (false- not trusted) information expanding at an incredible speed. By the time we detect a rumor in social media, it would have been retweed thousands of times; it would have become viral and would have caused undesirable effects. We would be lucky if we can still control the situation if a big disaster strikes, considering the speed of Twitter. Just remember the 6.000 tweet per second in the aftermath of Japan’s earthquake .That's why Twitter would the one to offer the Technological solution.
Using a HT helps, but it is no enough to stop rumors
What I describe above is a way to follow only the information of a topic and spare the rest. But, as explained, besides detecting the rumor stream there are some other issues to solve: where and when a tweet was posted (most tweets are not geolocated); whether the information delivered is updated, relevant, and trustful and who is the source of information (when this is not indicated). Our main problem remains. We need a lot of hands (VOST teams) to stop rumor spreading through social media.
As @LuisSerranoL says, one day we will face a mayor disaster amd rumors will be so big that we will not be capable of controlling them. Unless we find a way to really scrutinize social media and get the information delivered in a manner we can decide what is worth and what is rubbish at the speed needed to stop a mayor disaster.
500 million tweets per day and 241 million users deserve a solution to the Twitter conundrum, especially when we are talking of saving their lives, properties and the enviorenment. It is on Twitter's hands to provide a tool that helps emergency managers making civil protection to save lives using social media.
If Twitter succeeds in creating this rumor detector, the microblogging site would move from being just a mass media to effectively save lives in emergencies. This step would be the base above which Public Administration and VOST teams all over the world would build a holistic strategy to stop rumors.