The GovLab recently published a report on how social media analysis can be used to benefit a range of public good activities, from reducing traffic jams to increasing situational awareness in disasters. I have mixed feelings about the report.
Reading through the recommendations, I can only agree with the authors when they suggest concepts like the following:
“Social media corporations should consider themselves, and act as, the standard bearers for a new corporate paradigm of data stewardship. This coinage represents a move away from the concept of data as something to be owned and towards stewardship of data as a public good. (…) Data Stewards could help develop new coordinating mechanisms to unlock corporations’ supply of social media data sets with potential public interest value.”
However, I can’t help but feel that it is more than a little ironic that this would be in a report which was created with the support of Facebook, given that Facebook is famously reluctant to give researchers access to their data. It’s good that the authors appeal for more openness, but will this actually help change anything within Facebook?
On the other hand, it looks like Facebook is at least trying to find models for how they can share some of the data, while holding on to the data’s economic value, and I appreciate that this is not an easy task. The report lists some models that Facebook is exploring, such as sharing some data with trusted intermediaries, as is the case with Facebook’s new Disaster Maps.
In fact, the report does provide some interesting, additional information about Disaster Maps and also about how demographic data supplied by Facebook was used to improve a Zika campaign in Brazil.
Unfortunately, I find many other parts of the report disappointing. I’m always curious to see how social media can be used for qualitative analysis and here the report uses an example from 2012 when UNICEF looked at vaccination scepticism. While there is nothing wrong with that study, I would have liked to hear how Facebook or Twitter could improve on the methodology through their data for good initiatives, rather than reading about how UNICEF did it five years ago by themselves. Could Facebook make some of the non-public data available for studies like that? Does Twitter have suggestions how to analyse this more easily?
Similarly, I was a bit surprised to read about Google Flu Trends and ideas to use Twitter to predict the flu. These are valid examples, but dated. Much has already been written about them and I would have expected something that is a bit more current. The same goes for analysing historical Flickr data to see whether it can be used to predict flooding. While I am one of the few remaining committed Flickr users, even I have to admit that Flickr has become too niche to have any representative function. I fully support the idea, but why not repeat the experiment using Instagram? I’m guessing because Facebook-owned Instagram is more reluctant to share data.
And with that thought we are back to the question whether this report can actually move conversations regarding data stewardship forward, or whether funding research like this is just a way for Facebook to try to improve its tarnished image. I just hope that it does help to support those people within the company who are pushing for more openness.
As for the rest of us: the report really does have a lot of good recommendations and there is value in having these examples in one place. So take a look!
You can download “The Potential of Social Media Intelligence to Improve People’s Lives” here.