Slashdot Log In
Google Can Predict the Flu
Posted by
kdawson
on Tue Nov 11, 2008 07:42 PM
from the sees-you-when-you're-sleeping dept.
from the sees-you-when-you're-sleeping dept.
An anonymous reader mentions Google Flu Trends, a newly unveiled initiative of Google.org, Google's philanthropic arm. The claim is that this Web service, which aggregates search data to track outbreaks of influenza, can spot disease trends up to 2 weeks before Centers for Disease Control data can. The NYTimes writeup begins: "What if Google knew before anyone else that a fast-spreading flu outbreak was putting you at heightened risk of getting sick? And what if it could alert you, your doctor and your local public health officials before the muscle aches and chills kicked in? That, in essence, is the promise of Google Flu Trends, a new Web tool ... unveiled on Tuesday, right at the start of flu season in the US. Google Flu Trends is based on the simple idea that people who are feeling sick will tend to turn to the Web for information, typing things like 'flu symptoms; or 'muscle aches' into Google. The service tracks such queries and charts their ebb and flow, broken down by regions and states."
Related Stories
Submission: Google can predict the Flu by Anonymous Coward
[+]
Technology: Twitter Considered Harmful To Swine-Flu Panic 383 comments
judgecorp writes "Twitter is being criticized for spreading panic about swine flu. This is not just knee-jerk Luddism 2.0: it's argued that Twitter's structure encourages ill-informed repetition, with little room for context, while older Web media use their power for good — for instance Google's Flu Trends page (which we discussed last winter), and the introduction of a Google swine flu map." On a related note, reader NewtonsLaw suggests that it might be a good idea, epidemiologically speaking, to catch the flu now vs. later.
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Damn (Score:5, Interesting)
Thats a seriously great idea ...
It totally sidesteps the problem of early symptoms not typically getting people to the doctor where it can potentially be reported/tracked.
There's probably a lot of trends that can be detected the same way beyond just disease.
Re:Damn (Score:5, Funny)
I'd write more about why this idea won't work, but I'll have to do it tomorrow. Right now I've got a splitting headache, so I'm just going to put some neosporin on that bite I got from the weird guy on the subway train and then head to bed.
Parent
Re:Damn (Score:5, Funny)
You have to break your query down by language and how old it is. See, looking for "massive zombie outbreak" won't get any results if, say, Russia gets overrun by the undead. I mean, what's Russian for "Oh sh*t we're all gonna die!" anyway? And given how often this happens, you really need to sort by date too. I mean, two weeks ago there was a major zombie outbreak. It happened all over the country, like some kind of national holiday. And then the next day everyone was all like "nuhhh--what happened? Where's the aspirin? BrrrrAAAAaaaaIIIiinNNNnnsss" Damn zombie boys... get your own damn brains. *sigh*
Parent
Re:Damn (Score:5, Funny)
Yeah, because the first thing I am gonna think while running from a horde of zombies is, "Damn, I should go write a Slashdot journal entry about this".
Parent
Re:Damn (Score:5, Funny)
I fail to see the sarcasm in your comment.
Parent
Re:Damn (Score:5, Funny)
Yeah, because the first thing I am gonna think while running from a horde of zombies is, "Damn, I should go write a Slashdot journal entry about this".
Good point. But you can bet your ass that the twits on twitter will be tweeting...
"Zombies in the street. Gonna stay in tonight." ... ... ...
"Garbage stinks... better take it out."...
"it bit me. Hertz pretty bad."
"Man TV sucks on Monday night. Watching Simpons reruns."
"Seems cold in here. Crankin the heat."...
"I'm so hungry...lets see whats in the kitchen...!"
"Hand ii coodaafination fafading.. fafegae"...
"need bwrainsss brAaainzzs...."
Parent
Re: (Score:3, Funny)
"Hand ii coodaafination fafading.. fafegae"...
"need bwrainsss brAaainzzs...."
And these are different from normal twits (best ever name for twitter entries) how?
Re:Damn (Score:4, Informative)
what's Russian for "Oh sh*t we're all gonna die!" anyway?
According to Google, it's this [google.com]. It translates from Russian back into English as this [google.com], which reads as "Well crap, we all will die!" I'd paste it here, but we all know how Slashcode mangles any foreign languages or special symbols.
Oh, and I have no idea how I know this, but Russian for Brraaaaiiiiins is "Maaaassssgiiiiiiii".
So really, if Russian zombies ever invade, some poor sap is just going to think it's a hobo saying "Musky" and they get bit.
Parent
The Technique Works (Score:4, Interesting)
I posted models of it almost three years ago.
http://www.realmeme.com/Main/dailymeme/2005/Aug/coughcoldDejanews.png [realmeme.com]
Web searches are co-incidental indicators.
Want to see something that Google hasn't shown you?
http://www.realmeme.com/roller/page/realmeme/?entry=sars_versus_avian_flu_meme [realmeme.com]
It's quite likely that the Internet retains knowledge and alters its behavior over time. Compare the group reaction time between the SARS and avian flu viruses.
Parent
Re: (Score:3)
Or, at least, it worked great until everyone started typing in "flu outbreak" in google to find the page they created!
Second order effects, huzzah!
Re:Damn (Score:5, Informative)
Parent
Re: (Score:3, Insightful)
Well, if you RTFA, you'll see that Google's method applied to the past four years very closely mathches trend data collected by physicians in coordination with the CDC. The proof is in the pudding
Exactly, which is why I'll be impressed when they can do this ahead of time. I'm not holding my breath. Analysing data trends in existing data and concluding you can predict them is not impressive.
Re:Damn (Score:5, Insightful)
That's why we have a "flu season." It is very cyclic in nature. Trends will very likely be a good indicator when localized spikes of new queries provide a precursor where a previous trend can further enforce.
In other words, a spike of localized searches related to flu falling well within flu season for a given geographic locale, likely is a precursor to a growing flu outbreak. It's really not that hard to imagine - especially once you consider the incubation time of your typical flu virus.
The lead time prediction of Google's method verses the CDC's post-suffering reporting is easy to guess. The CDC's numbers measure reported cases. Google's method measures localized interest (signal), develop a metric to discern against baseline interest (noise), and apply against trend data (signal has velocity), you likely have identified a growing flu outbreak. Once you add the incubation time, it's likely Google's numbers have a strong correlation with the reported CDC numbers.
Parent
Re:Damn (Score:5, Funny)
The proof is not, nor has it ever been, in the pudding. However perhaps you meant to say the proof (i.e. test) of the pudding is in the eating.
Idiom police at your service. No, you needn't thank me ... just doing my job.
Parent
Re:Damn (Score:5, Insightful)
It doesn't predict anything reliably. Too many variables.
Simply put: If you're looking for help online for flu symptoms, that doesn't correlate with an 'outbreak' of flu.
And what defines outbreak anyway?
Well, the way flu works, if you have it, you're likely to give it to someone else. You may google about it when you don't actually have it, but how often does that happen? The number of false positive searches would probably be somewhat low, and either way they would be constant. Google serves millions of search results a day, if not more. Almost everything "random" would, over time, look constant. When non-random things happen, like people from a certain region (remember, google knows your IP) getting the flu, even a 1% increase in flu related searches is extremely significant, if it otherwise doesn't vary that much.
YOU googling for flu symptoms doesn't necessarily indicate if you have the flu, but a large increase in the number of people googling it probably does. Especially if you can compare your data to the CDC data, to check your theories.
-Taylor
Parent
Re:Damn (Score:4, Insightful)
If many, many others in your area are doing the same, it just might indicate a local outbreak.
Graph this over time, and you might see trends happening.
Do this for a couple of years, and compare to actual CDC data, and you might just find it works.
Parent
Re:Damn (Score:4, Insightful)
Simply put: If you're looking for help online for flu symptoms, that doesn't correlate with an 'outbreak' of flu.
I think you need to look up 'correlate' in a dictionary, you obviously have no idea what the word means. A correlation is not a one to one relation, if A correlates with B all that means is that A is more likely if B is true.
Sure, the fact that i just went and searched for flu stuff out of curiosity doesn't mean there's an outbreak near me, but people presumably perform searches on this at a pretty steady rate, and a flu outbreak ought to cause a spike in searches. The occasional false positive happen in a region, say if there's a news story on the flu, but to say there's no correlation is ridiculous.
You could I suppose argue that the correlation is too weak to pick out from the noise, however if you RTFA, it is quite clear that the correlation is quite strong enough to produce useful results.
Parent
Re:Damn (Score:5, Interesting)
What we are seeing is the edge of a very powerful, useful, and potentially deadly technology.Given a large enough quantity of data it is quite likely that correlations of seemingly unrelated data can be used as accurate predictors for other events. For example the price of eggs might be used to predict movement in the price of gold. Obviously that is overly simple but by using computational power and large data bases there should be methods found of predicting all kinds of things that may be world changing.
Parent
Re:Damn (Score:5, Interesting)
Not according to the graph here [google.org]
Parent
Re: (Score:3, Informative)
If you watched the animation you would realize that the "general depiction of a rising and falling curve" is the point. The google prediction is two weeks ahead of the CDC data for the same changes, and can be data mined far more specifically for location and such.
Re: (Score:3, Insightful)
The CDC charts reported flu cases. Flu cases are only reported if you seek medical care. If you just go to the drug store and buy a bottle of NyQuil, the CDC doesn't know you had the flu. I should not have to tell you this...
So, if you want to game the system... (Score:5, Insightful)
Buy stock in companies that sell treatments for Beri-Beri, Trench Foot, and Jungle Rot, and then have your botnet look them on on google.
-jcr
Re: (Score:3, Interesting)
Buy stock in companies that sell treatments for Beri-Beri, Trench Foot, and Jungle Rot, and then have your botnet look them on on google.
I had thoughts along the same line, but in a different direction after reading this quote FTFA:
"And internally [Google] has tested the use of search data to reach conclusions about economic, marketing and entertainment trends."
If Google can correlate search terms with movements in stock or commodities markets, they might be able to attain that precious first mover advantage [wikipedia.org]
future Google services (Score:5, Funny)
This, of course, won't work on female users since we all know that girls don't poop.
Re:future Google services (Score:5, Funny)
yes we do, we just don't make a damn production out of it by telling everyone, then doing it and acting like setting off the smoke alarm gives us extra credit.
Parent
Re:future Google services (Score:5, Funny)
Parent
Madagascar Closes Down (Score:5, Funny)
There are other possibilities (Score:2)
Can slashdot cause flu? (Score:5, Insightful)
Do to the /. effect thousands of /.ers started googling flu symptoms causing the predictor to indicate a flu outbreak.
Thousands of hypochondriacs responded by checking themselves into hospitals complaining about flu-like symptoms.
Finally, a get rich scheme that will work! (Score:5, Interesting)
1. Invest in Tamiflu (the leading medication to treat flu symptoms)
2. Organize a massive effort to do web searches for "flu symptoms"
3. Wait for Google to sound the flu alarm
4. Profit!
Trend data as a long term resource (Score:5, Insightful)
As a capitalist, and an incubator, I've spent tens of thousands of dollars (per project) on market analyses. For me, finding if a particular good or service, even a niche or very specific on, is desired in a given area is expensive. It's often the MOST expensive thing I do before starting a business.
I've always harbored the idea that Google's grasp of data, even just raw data, is their most important resource. As they make this information available, the market will prosper. I've been able to use Google Trends (national, not local) to profit from the so-called "long-tail" and enter a business market I might otherwise not have.
When Google starts making trend data available based on region, it will be a huge boon for guys like me -- the risk takers. I'd love to know if a certain term is growing in popularity in given regions, or even in given regions at certain times (say "Where can I get vegan food?" in Chicago after 10pm but before 4am). I'd love to know if it's from a desktop or mobile, or even a Mac versus PC. By digging deeper into a customer-base's desire, Google trending can offer me a profitable business, but it can also offer the customer base more competition (or even a product that isn't readily available in their market).
The flu trending is just an eyewash to push Google's strength in raw data retention over time. That's their reason for doing it. Will it help people? Certainly. But to those anti-capitalists, this is exactly where capitalism reaches those in need, but still can provide a profit for the charitable person or company.
Re: (Score:3, Insightful)
I had a similar idea.... (Score:4, Interesting)
We all know that animals act odd, hours or days before things like earthquakes. The morning before a 6.8 quake in Washington State in 2001, my neighbors dog that normally will do anything to force it's way out of the front door and run for hours when the door is opened the slightest bit, wouldn't even get close to the door when the neighbor opened it. By itself, 1 animal acting weird means nothing, but a large group of animals over a localized area acting weird at the same time would point to something about to happen. The problem is that it is always after the "catastrophe" that people say, "You know sparky was acting odd this morning". If there was an online database that you could quickly go to and report that at your address your pet is acting weird at this moment, you might be able to predict the event by looking for groups of "odd acting" pets. I know it seems like a weird & far fetched idea, but tell me why it wouldn't work.
Re:I had a similar idea.... (Score:4, Insightful)
Parent
Guys named, Dick Drop. (Score:3, Funny)
"That, in essence, is the promise of Google Flu Trends, a new Web tool... unveiled on Tuesday, right at the start of flu season in the US. Google Flu Trends is based on the simple idea that people who are feeling sick will tend to turn to the Web for information, typing things like 'flu symptoms; or 'muscle aches' into Google. The service tracks such queries and charts their ebb and flow, broken down by regions and states.""
Hmmm. *types in Google "Dick falling off"*
Self defeating? (Score:4, Interesting)
Syndromic Surveilance (Score:3, Insightful)
This sort of thing has been floated around for a while under the banner of 'syndromic surveilance'. I spent most of the last three years working on a research project that involved gathering data on water quality and developing statistical software to find subtle indications of contamination. The intent was always to extend the approach to syndromic data, incorporating things like over-the-counter medicine sales, ER visits, and so forth.
Unfortunately, it turns out that none of us on the team knew enough about statistics to manage a fantasy football league. I'm now happily self-employed doing stuff absolutely unrelated to statistics. I think some of my hair has grown back, and I hardly even cringe when someone says 'generalized least squares'.
If you're interested, though, here [niusrjournal.org] is a paper from the CDC on the subject. I'm pretty sure they have a better idea what they're talking about. Or at any rate, they've got nicer graphics.
Re:Great. (Score:5, Informative)
If you've received a flu shot in the past 6 years the only thing you got was a chance at a bad immune reaction and a concoction of mercury, detergent and some other nasty compounds.
That's blatantly untrue. The flu strain predictions have been fairly good in the last few years, with the exception of 2003-2004 when it was only marginally protective for one of the more common strains. Even that year, it was largely protective for most strains. Get your damn flu shot and protect the rest of us. For reference:
wikipedia [wikipedia.org]
Parent
Re:Great. (Score:4, Interesting)
A realistic assessment of the flu vaccine can pretty easily show its value- it's around 60-70% effective, according to the sheet they gave me when I got vaccinated this year. If a majority of the people you come in contact with are vaccinated, it clearly reduces the probability of infection. This becomes especially important if you plan to visit anyone in a nursing home or hospital, in terms of protecting them as well as yourself from the flu.
So stop trying to out-think the logic of vaccination just to be different and go get vaccinated. It won't hurt you, it doesn't cause autism, and you won't turn into a zombie (and even if you do, brains may just be pretty tasty)
Parent
Re:Great. (Score:5, Informative)
Umm...
You're TOTALLY wrong. WHO-recommended flu vaccines are very effective. See here for an example: http://www.phac-aspc.gc.ca/publicat/ccdr-rmtc/06vol32/acs-07/index.html [phac-aspc.gc.ca]
And: "...the only thing you got was a chance at a bad immune reaction and a concoction of mercury, detergent and some other nasty compounds..." is just a stock anti-vaccination quackery.
Parent
Re: (Score:3, Insightful)
How about all the people who haven't gotten s
Re:Great. (Score:4, Informative)
If you've received a flu shot in the past 6 years the only thing you got was a chance at a bad immune reaction and a concoction of mercury, detergent and some other nasty compounds.
Thimerosal (mercury) is only used in multi-dose vials. Although these are legal in the US, they are in practice not used here. The chances you received any thimerosal in your flu shot if you got in the US is almost nil.
I do agree with the OP that two years of the last decade the WHO predicted which strains would be dominant in the US incorrectly and thus the shot didn't immunize the recipient properly against the strains they would actually face.
Parent
Re:Great. (Score:4, Interesting)
Just an aside.
How the hell does WHO predict Flu strains for immunization? I am honestly ignorant and would like to know.
Parent
Re:Great. (Score:4, Interesting)
Influenza (the flu), originates in China pretty much every year. Different strains of RNA viruses arise (mutations in antigens, the process is called antigenic drift), in pigs. These are different enough to be able to get past the immune system (which is resistant to last year's strain, but unable to recognize recombined (new) antigens).
Pigs share a common receptor with people which is hijacked by these RNA viruses; this allows the virus to jump from pigs to humans. (Aside: they also share a different receptor with birds, which is why we're so paranoid about avian influenza: it could jump to pigs, mutate to our receptor, and then jump to humans.)
Every year you'll have dozens of different strains of influenza arising in pigs; only and handful of these use the common receptor and are able to jump to people. From there, only a handful of these are spread (through migration) to other parts of the world. IIRC, the flu spreads west with the climate, eventually encountering a city where it's able to hitch a ride to America (and the rest of the world) on a boat or an airplane.
WHO relies on being able to look at previous strains which reached epidemic and pandemic proportions, and on being able to artificially recombine antigens to create this years major strains. Sometimes they miss a few critical ones (2003-2004), but they're remarkably good about predicting which strains will mutate.
Basically, there are only a few different antigens, and we rely on creating the same new set that nature will create (there's a finite number of viable recombinant strains, after all). I doubt they look at pigs in China; there's simply too many in areas that are too remote.
If there's an epidemiologist reading this, he can probably give you a more detailed answer.
Parent
Re: (Score:3, Informative)
How the hell does WHO predict Flu strains for immunization? I am honestly ignorant and would like to know.
No magic, really.
Basically, they pick the dominant strain that is circulating at the time that they have to make a recommendation to the vaccine producers.
The vaccine strain has to be picked ~6 months before it is needed (it takes that long to grow it up in eggs in sufficient quantity). Typically, that is right at the peak of flu season for the other hemisphere (north/south).
The selection is based on the RNA sequence for the virus, and on antigenic tests (antibodies to the strain, grown up in ferrets usual
Re:How long until... (Score:4, Insightful)
Parent
Re:So? (Score:5, Informative)
Parent
Re: (Score:3, Informative)
Re: (Score:3, Funny)
everyone knows it starts in arnette, tx.
Re: (Score:3, Interesting)
Yes, but not without violating the privacy of the users of google. We don't ask for their personal information from the clinic because it's a privacy violation, but we apparently have no moral objections to using their browser histories to get pretty much the same data.