Tracking the Flu with Twitter

According to a CNN article published a couple of weeks ago, researchers are trying to use real time tweets and Facebook posts to map out potential and/or current flu outbreaks. Researchers at Johns Hopkins University have found a new method for filtering tweets which could make the real-time data pouring in more accurate by using special algorithms, that are able to filter in words like sick, flu, headache, etc.

However, there is room for development. It is very common for the algorithms to associate wrong tweets for instance if it says, “my new car is sick” or “Bieber fever“. Nevertheless, researchers are trying to improve its accuracy and even predict when and where will the next outbreak be. This kind of predictions could be vital for hospitals and clinics as well as for regular people, as a way to prevent and handle situations like getting a flu shot before your community has an outbreak or having enough doctors and  nurse staff.

Furthermore, the article also mentions a couple of other apps and search engines such as Google and Germ Tracker who are designed and are helping track and predict future outbreaks. The CDC (Center for Disease Control) also acknowledges that Google’s findings are almost always in line with the CDC’s predictions. This could be another incredible use of social media and help organizations like the CDC to a better job of protecting communities.

My questions for the class are:

1. Can this data be trusted? Don’t we often over exaggerate what is going on in our lives? How accurate can it be?

2. Can this be the beginning of a new liaison between Social Media and the public health industry? Could we predict other things besides illnesses?


3 thoughts on “Tracking the Flu with Twitter

  1. Yes, I think that as the technology improves it can be a valuable tool to help medical professionals add a “social lens” that verifies or strengthens their other findings.

    I am sure this sort of social listening can be used for other predictive evaluations ranging from politics to crime.

  2. I feel like this kind of data cannot be trusted. I agree that we often do over-exaggerate how we feel on social media platforms in the hopes that people will feel bad for us. I also think that social media is an unreliable source in general. Just look at the catfish phenomena, who is to say that there aren’t fake Twitter accounts as well. I don’t think that this will catch on and be used as a reliable illness or epidemic forecaster

  3. I love that this exists. I think it’s a really smart idea, even though there are obvious glitches as you mentioned. This definitely seems like something that could be really, really useful (even on a global scale), although it would take some serious tweaking. Although, given that CDC claims that Google is accurate in its germ findings, we could be well on our way in terms of developing this type of program. Maybe it’d be better to create a search for common phrases like “not feeling well” versus particular words like “fever” or “chill” that could easily be applied to any number of instances. Or maybe there’s a way to geo-fence tweets to particular locations (not really sure how to do this), so that if there’s a germ outbreak near you, you’ll know about it, without having to know about the rest of the germs all over the country.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s