Think about where you are right now… How would you describe your location?
In the Board Room at Sequoia Capital.
At Rockefeller Center.
On the Hollywood Walk of Fame, near Hollywood & Vine.
In the Glamour Shots store, on level two, within Mall of America.
In Casino Niagra, at Niagra Falls.
At 301 Front Street West, in Toronto, Ontario, Canada (which most local Torontonian’s don’t even recognize as the address of the world’s tallest free standing structure – The CN Tower).
Invariably, people describe their location in reference to a place. I have yet to have anyone respond to this question with a lat/lon coordinate (perhaps in a military operation this might happen). Thus the vast majority of digital information we have created or collected is associated with places, and not currently encoded with geographic location.
Now, imagine trying to use the internet to find all the information, nicely organized, about the other places within a mile or a block of where you are. I am not talking about map-enabled yellow pages or sketches of buildings in Google Earth… but rather all the content in the web, as it relates to the places nearby. Phones, PDAs and personal navigation devices like TomTom are attempting to ley you search by Place. The cool new Blackberry Curve 8310 is almost there now (and if rumours are true, perhaps a future version of the iPhone). GPS positioning and map display on wireless devices are fast becoming ubiquitous. The users now expect that all the content about places of interest is there as well, and searchable by location.
Consider the responses above. There are important relationships between and among these places – a real world hierarchy if you will, which we can traverse quickly in our minds due to spatial intelligence and geographic knowledge. Anyone that has taken a Geography class or ever studied a globe, is able to create a mental picture of roughly where these places are in the world. Whereas, most computer systems simply don’t understand the geographic relationships inherent in the data they manage, even with the advent of and exploding interest in Google Maps.
Indexing information for local or geographic search, and linking it to a map is still a very gnarly problem. Even John Hanke, Google’s mapping king admits that currently there is limited content linked to their maps, and “searchable” by location – he goes on to say that as the amount of content grows, it will require a new method or technology to “sift the wheat from the chaff”. We believe this sets up the value proposition for a universal Place ID scheme. Unique ID codes or numbers already exist for books, cars, people, and a myriad of other entities – but not for places.
This content organization challenge will not be solved by geocode/geotag or mapping technology. Rather, it’s about resolving and correlating varied and often ambiguous place references, and understanding the hierarchy of say a store, to level 3 of the mall (which in the case of Mall of America, has 9 valid street addresses and associated geocodes). It’s about place – not location. This is where the Place ID comes in – it eliminates ambiguity and provides precision down to sub-building level, for data indexing and integration purposes. Putting a pin in a map, to show (roughly) where that place is in the world, is relatively easy – in fact it’s pretty much a commodity today.
Getting back from a lat/lon position to real world place, and all the related content, simply can’t be done today.
A lat/lon pair can’t serve as a unique Place ID because with 20 – 30 meter error factors, it is not precise enough. As well, addresses fall short for this purpose, due to a lack of standard formats, as well as frequent errors or discrepancies. Many places of interest, such as the Grand Canyon, don’t even have an address that can be geocoded. Cquay’s Place ID fuses place name, address, and location (lat/lon point or boundary defining a geographic area) in a single standardized key. With this key, content about places can be tagged or linked to a master index, and aggregated with high precision and zero ambiguity. It is also granular enough to be the integration key for brand, product category, and inventory associated with a particular store within a mall.
As context and localization become increasingly important to the user search experience, places are becoming the new keywords used in a search bar. Behind the cool map-enabled search interface, a next generation information integration and search technology is needed. In 2001, we coined the term location intelligence to define our vision for a place-centric search method and platform, called Common Ground(r) that would link a world of information about places -to web-based maps. Today, the value of this technology and an associated market is starting to emerge in a significant way.
To be continued…