Floogle

[bholus10]

A lot of people are unaware of Google.org, the philanthropic wing of Google. Since its inception, Google.org has been involved in a variety of initiatives covering such global issues as health and disease, climate change, education, poverty and alternative energy.

Their involvement includes grants for a wide variety of projects, encouraging employee involvement and use of their own innovative technologies to help make the world a better place.

While this program has its detractors, there are those who claim Google does this only to toot its own horn or to make money. While this may be true, Google does seem to do quite a bit more than most, if not any other corporation, to help improve the world we live in.

Last week I was reading about Google.org's latest venture, Flu Trends, which I found to be a very interesting use of their search technology.

Basically, through coordination with experts in the field at the Centers for Disease Control and Prevention (CDC) and using search data on specific "flu related terms" they have started to track overall patterns of flu infections throughout the US.

While there is some question as to how accurate using search terms is to determine how many people are actually sick in a given area, the results are surprisingly close to the data collected by the CDC. Also of interest is that this data was obtained up to 2 weeks more quickly.

The use of search trends is hardly new; Google has this incorporated into its keyword research tool to help track trends for marketing and keyword visibility purposes.

[bholus10]

It also has Google Trends for viewing and comparing specific keyword search patterns graphically.



Google Insights for Search can also be used to obtain this sort of data categorically, by date range and across specified regions worldwide.



This provides a nice map interface that can be viewed from a world level right down to a state/regional view.

State View

[bholus10]

So, essentially, all Google has done with Flu Trends, is to take data anyone can get using their tools and focused it on one specific set of search terms and made it available for view on a site for everyone to see. There are some differences though.

Flu Activity New Mexico



Flu Activity Mississippi

[bholus10]

In Flu Trends, you can click from state to state to compare and also you can see results from previous years charted in advance of the current year's display progress.

The thing I find odd, is the geographical limitations they've imposed on the Flu Trends site. They could easily display the same type of information for countries other than just the US. Nor do I really understand why it's limited to state level. They could just as easily show flu trends by city.

Google can track these trends by geographical region because each search query is tagged with the searcher's unique IP Address. Anyone can track the rough location using software or even web based applications. As an experiment I used the latitude and longitude values for our office IP address that I obtained here, pasted them in the decimal value fields on this site and a map location within about 9 blocks showed up.

Certainly not exact, but close enough to be able to track search queries at least by city, if not down to individual municipalities. Personally, if I was after this information, I would rather know if there was a big surge in my immediate area, rather than the whole state.

Maybe they don't want to make users nervous about privacy issues, as seems to be implied in this interview with Elizabeth Landau, from CNN.com.


As Google has IP's for all search queries, I can't see the fuss about them knowing who searched for flu related terms, but that's just me.

While this seems like a great idea, there are shortcomings with this sort of tool, for example:

People who are sick with the flu but don't search.
Searchers not actually sick with the flu, just doing research for whatever reason.
Searchers looking up info out of concern for a friend or relative who is sick but in an entirely different area.

I'm sure this is just a first step as there's much, much more that this could be used for, as we can see on the Global Health Watch site (incidentally powered by the Google Maps engine), the incidence of influenza in the US is not the only going health concern.s

[bholus10]

What else could they do?

Google is uniquely placed for these types of services; not only because they have the resources and the apparent desire, but because of exposure. A lot of people go to Google, not just for search, but for their mail service and a host of other integrated free services.

Given that, I starting wondering what else they could do to improve the accuracy and utility of this sort of thing.

One way I could imagine, would be to have some sort of direct user input - something like a Google suggest tool for self diagnosis.

I can see how doctors and lawyers both would be horrified at what I just suggested, but I'm talking about a way to not only track patterns of health issues but also to better prepare the consumer who is ill for the next stage(s) of dealing with their issues.

Obviously, this is not to provide a substitute for consulting a physician.

I know I tend to drag my heels about going to see a physician unless I'm bleeding to death, but by the time I get there I forget a lot of possibly important details that my GP might not be able to dredge up in our brief appointment.

I would love to have something like a logical, online questionnaire/checklist with some sort of flowchart type progression of questions based on my answers that I could fill in at my leisure.

Even better, if I could then print off the results and hand it to my doctor when I got to my appointment. This would save their time and reduce the number of irrelevant questions that need to be asked.

Even if the tool only gathered symptomatic data, rather than conclusive diagnostics, it would still be valuable. For example: if in a short period of time an specific area shows a higher than normal set of specific symptoms,

it could tip off health authorities or even people browsing the data that something out of the ordinary is taking place and they could track down that contaminated water supply, or previously unknown environment hazard or whatever.

Likely this could enable problem identification and resolution much quicker than by normal means.

Another public service field that Google could provide is a plug for the huge gap in emergency notification.

Currently, disaster notification is kind of all over the place. There are all manner of solutions ranging from those that require people to have their TV or radio on right through to commercially available corporate notification and disaster recovery services, and anything and everything in between.

One great model for updated alerts and information is GDACS, but that is geared towards keeping the international emergency response community notified and not designed to handle specific area notification to the average Joe.