I knew it! The state of POI data does suck!
So I can still recall driving around Manhattan with my brand new Garmin device circa 2005 just playing with all the cool features and seeing what I could see… it certainly didn’t work perfect… particularly living in midtown where all those tall buildings make getting a GPS fix difficult, and where it could very easily show you a block or two off on either side, making things even a bit more confusing… something to do with signals bouncing off buildings I think.
But what resonated with me most, was chuckling at the business listings that were purported to be surrounding me as I drove down fifth avenue on to Central Park South… home to some of the toniest hotels and shops like The Plaza and Pierre Hotels. According to my Nuvi right next door to those hotels was supposed to be a place called AAA Als Towing and then a few blocks later an auto repair place… yeah right, how many auto repair and tow places do you know paying more than a few grand per square foot for such prime real estate… it was apparently the early days of POI spam, or at least a really bad dataset.
You’ve probably seen articles about the locksmith map spam problem on places like Google Maps, but its much more than that… the industry as a whole suffers from just really poor information related to documenting places… Read more
You Deserve a Rake Today at Ickdonalds
Some interesting posts over at Blumenthals.com (currently one of my new favorite blogs along with Maperture) about Google’s decision to allow anyone to modify Google Maps placemarkers and the information contained within, until the listing is claimed by the true listing ‘owner’. It turns out that disgruntled customers (or employees) are taking the opportunity to disparage businesses that have wronged them by taking over their listing and doing not so nice things to it… perhaps you’ll really enjoy the fries from that Ickdonalds, even if it is strangely situated in the middle of Lake Erie!
It does bring up some interesting issues like why in the world would Google not regulate this more closely and why they feel the need to allow users to have free reign over a core part of a Google product… even when it has the potential to erode the quality of the product.
I think the answer is tied to both the economic and ideals over at Google.
While Google of course has more money than God and can and does spend money easily without a clear path to a return on investment, the time and effort it would take to develop a monitoring and filtering technology to handle this is not insignificant, particularly for something like Google Maps which is probably not yet a significant contributor of revenue. In fact Google set this precedent as the preferred modus operandi back when they acquired YouTube and opted at first just to take down copyright infringing content rather than trying to filter videos as they went up. The Viacom suit of course changed that.
What it also seems to be saying is that the current state of POI map data is not where it needs to be. Afterall, if everything was accurate and up to date, why would you need to leverage the crowdsourcing in the first place. And if Google WERE to build or license something similar to what they used to filter for copyright infringement for YouTube, they’d need a “master copy” of the “correct” data against which to cross reference… which of course doesn’t exist and would be a full time job to create and maintain… ask Navteq… and besides if they had that in the first place, there would be no need to crowdsource it and the issue would be moot.
The other part of the answer is that Google is not a content company… ie their mission “is to organize the world’s information and make it universally accessible” … so nothing in there about creating original information itself or ensuring the quality of the information, they just help you find it… no matter how much it may suck. So the ideal in Googleland would likely be that all the POIs just existed out there in the world and they would happily make a copy, store, index and help you find the stuff you’re looking for and be on their merry way.
But alas, all the worlds information doesn’t always cooperate they way you’d sometimes like.
Content is king, long live Dogparkusa
If you’ve worked in the media industry for any amount of time, you’ve surely heard folks declare “Content is King” more times than you’d care to remember. If you’ve worked in the web world for any amount of time, you’ve also heard your fair share of dissenters declare that if indeed Content is King, then the King’s castle is currently being rushed by a mob of peasants, in the process of breaking down the front gate, looking for their own share of the power, spotlight and good life.
The age of community created content has arrived online and is slowly disrupting everything that old media once held sacred. In the media landscape of a decade or two ago, the creator of great content could make one hell of a connection… who the heck remembers what kind of television they used or what network carried the programming when they watched the last episode of MASH, or Cheers, or their favorite team in the SuperBowl? Uh, no one, but I am sure the content itself made quite an impression. But back then you could also count the number of media outlets on your fingers, and the influence of each was magnified by the scarcity of alternatives.
Up until recently, mobile map content was likely supplied by one of two companies: NAVTEQ or TeleAtlas… not unlike the ABC, CBS and NBC of old. The two companies sat back and decided what they thought the world needed in their maps, invested a bunch of money to go gather that data and periodically publish up the results. Repeating as often as was economically viable and demanded by their customers. Not much different than ABC deciding what the world wanted/needed to watch, and going out and buying shows that appealed to the broadest audience possible and offering up 22 episodes at a pop.
I suspect that we may see the same occur with location data content. While we have seen innovative initiatives like community editing of map data by TeleAtlas, today this is still just focused on the basic blocking and tackling of getting the base map data updated and correct. In order to provide much more rich information about the world, niche specialist will need to step in and become data providers for their areas of expertise… for example the Travel Channel can step in and become the leading provider of information related to popular destinations for tourism, news organizations like CNN will become providers of breaking timely news information related to places, and similar roles for Zagat or Michelin in restaurants, AAA for travel essentials, etc.
At this stage in the game, there may not be enough mobile location aware users to support the next phase of micro niche content creators purely for the mobile location experience, which is why you see folks trying a lot of different approaches with user generated content… either just opening it up to anyone creating content about anyplace, like Platial Nearby (the GeoCities approach), or by trying to limit to content created by someone within your social network, like Whrrl (the Facebook approach), with many trying to offer a blend of both alternatives. Alternatively, you have folks like uLocate not creating content at all, but aggregating what it sees as the “best of” in content and applications not unlike the original directory/portal set up of the early years of Yahoo!
But it’s looking like it won’t be a big leap for many of the niche content creators that have emerged on the web to transition over … take for example the husband and wife team at dogparkusa.com… they along with the community they’ve built, have aggregated information on over 1,600 dog parks across the US. They’ve already done a great job of integrating a mapping element into their website and even use Lat49/Local.com to monetize those web map views. But do you think you’ll find their information when you’re driving around with your latest TomTom or Garmin or Nokia device, or even in one of a myriad of mobile finding apps like Earthcomber, Where or Whrll? No, Not Yet… and I don’t suspect that you will until the mobile location aware equivalent of AdSense/AdWords is invented to make it profitable for both sides to do so.
Centralized versus Edge Storage For POI information, the Akamai Model
When you hire Akamai you’re generally looking to 1. ensure that a user anywhere in the world can load your website quickly while 2. avoiding the unnecessary expense of buying extra servers to be able to handle rare peak traffic periods and a geographically diversified user base.
Akamai provides their service by storing files that make up your web site out on the edges of the network, not one time at a single central server, but with many copies of your site on many severs physically close to the user, maybe just a few miles from your house or office, to minimize network traffic and the latency, problems and congestion that ensues. That sounds like a herculean task, until you learn that approximately 80% of the U.S. population lives in urban areas which make up just 3% of the land area… then it seems a little more doable, although still impressive.
Knowing how Akamai works and the correlation between physical proximity and performance, makes for an interesting questions about the best way to communicate location data about points of interest (POIs) to the nearby users of mobile device that are location aware.
Let me first clarify, err steal from Wikipedia, what the heck a POI is, from Wikipedia it’s “a specific point location that someone may find useful or interesting” . Most people think of things that are always there like businesses, parks and government building, etc that are useful to everyone when talking about POIs. But in my mind POIs also include locations that may be of interest to a select few like a geocache, or only of interest for a two hour window because an event may be happening there, or even just for an instant because some thing of interest passed by that point.
So with POIs that don’t move, are ‘always on’, and are applicable to everyone, here is how it works currently. Let’s say that there is a Wendy’s at the corner of Main St and Maple Avenue, over time the presence of that store gets recorded in the various databases of folks like Acxiom and InfoUSA and then into Navteq, TeleAtlas and on into Google Maps, Y! Maps, and Virtual Earth as well as Garmin and TomTom PNDs for consumers to find. And if the Wendy’s closes or moves, slowly over time the old listing will make its way out of the system after someone drives by or tries to call Wendy’s and notices it is no longer there and updates the database. The biggest problem with the current system for these types of POI’s, is that the information is often static and stale… gathered by an outside observer at a point in time, typically way in the past.
The biggest problem with the current centralized storage system for all other types of POI’s is that it just flat doesn’t work.
So what is the solution? Well what about storing POI location information locally, out at the “edge”, in mobile location aware units at the physical point of interest that can wirelessly broadcast out their presence and details about what they are (a store, event, a bus, a hotdog cart) etc to other nearby mobile location aware devices (ie phones) and cut out the middle man and all the headaches in the process.
I am sure there are a thousand reasons why it would never work, but heck I’d love to start hearing some of them.
-
Sponsored Links
