The Case For Traveling to the Center of Our Social Networks

James FowlerMuch has been made of David Lazer’s finding that Google’s Flu Trends tracker seriously missed the mark in its measurement of flu activity for 2012-2013—and in previous years, too.

For those who don’t know, Flu Trends monitors Google search behaviors to identify regions where searches related to flu-like symptoms are spiking.In spite of Flu Trend’s notable misstep, Lazer still believes in the power of marrying health and social data.

In discussing the results of his study, he has maintained Google Flu is “a terrific” idea—one that just needs some refining. I agree.And, earlier this month, Nicholas Christakis, several other colleagues, and I—with funding from the Robert Wood Johnson Foundation—published a new method offering one such refinement.

Our paper shows that, in a given social network (in this study’s case, Twitter), a sample of its most connected, central individuals can hold significant predictive power.

We call this potentially powerful group of individuals a “sensor group.”

By finding and monitoring the tweets of a sensor group, we can catch—and sometimes even predict—the outbreak of contagious information early on. That detection edge could improve how we track the outbreak of disease epidemics, the rise of certain terms or phrases, or shifts in political sentiment.

Whereas Flu Trends relies on a relatively static, proprietary “dictionary” of flu-related search terms based on average Google search habits, the sensor method taps into what is really happening in social networks in real time.

By drawing on language being used by a sensor group—such as mentions of an emergent symptom or a popular newly coined name for a disease—Google could gain insight into what their dictionary might be missing.Sampling both the average Googler’s behavior and that of the exceptionally connected social network user can paint a much fuller picture of whatever landscape we are interested in tracking. We can more accurately see how it looks now—and how it could look in the near future.

Our method could have implications well beyond Flu Trends, improving how we use social data for a variety of public health monitoring and intervention efforts. For example, imagine if we started talking about health issues  using the phrases sensor groups indicated real people are using—or about to start using.

We could proactively customize and continue to evolve health education campaigns to ensure they remain resonant with different communities or regions.There is more work to be done, but I am optimistic social data will continue to provide us with smart ways to improve the health of individuals and our communities.

James Fowler is Professor of Medical Genetics and Political Science at the University of California, San Diego.  He recently co-authored a book on social networks, Connected: The Surprising Power of Our Social Networks, with Nicholas Christakis. This post originally appeared in the RWJF Pioneering Ideas Blog

2 replies »

  1. I love this idea. I too, feel Google Flu Trends has huge promise. But I wasn’t overly surprised by the recent setback. From what I understand, this is a sensible compromise. If there is any takeaway from GFT it is that data needs to be filtered to be understood, like any other form of information. We have to resist the temptation to do too much to soon – or better yet, do too much by all means, but remember that we’re pushing off the edge of the map …

  2. It’s funny that people were surprised Google’s Flu Trends were found to be off. I suspect that as more and more of the these predictions are made with big data, more and more cases of them being wrong will make the news. Even if a prediction algorithm is 80% correct, that leaves 20% incorrect predictions. All of those incorrect cases will start to add up to many wrong predictions.

    It will become more like the weather; we’ll get used to all of the wrong predictions and start to look out the window.