Exploring Local
Mike Dobson of TeleMapics on Local Search and All Things Geospatial

Gate Keepers – Part 7 – The End

April 2nd, 2009 by MDob

The hybrid approach to creating map databases, which I define as the coupling of a defined, rigorous map compilation process with user generated or crowdsourced map data, seems to me to be the future of commercial cartography. While some would argue that crowdsourced map data will replace all other types within a few years, I don’t see that happing.

Crowdsourced map data is of good enough quality for some uses, but not for others. I doubt that you will see ADAS applications based on unverified, crowd-sourced data. In addition, I doubt that the measurement devices needed to create precise ADAS certified map data will be in the hands of “everyman” anytime soon. Next, it is unlikely that government agencies responsible for providing maps that meet “National Map Accuracy Standards” will relinquish the “crown jewels” to “Joe Sixpack” for maintenance, augmentation and evaluation.

However, in many applications the hybrid approach will produce “fresher”, more accurate and more widespread cartographic data than can be gathered by other methods. In fact, neogeographers and neocartographers are assuming the roles that used to be the domains of academic geographers and commercial map publishers. Maps, map data and geographical analyses are being published by concerned citizens who have now found a way to tell the spatial stories that are of interest to them. When they find the map data lacking, they will attempt to augment it themselves, or look to someone else who is concerned about the same problem for a solution.

Unfortunately, motivation to tell a story does not necessarily equate with understanding how to use spatial analysis and cartographic process to tell the story in as unbiased a manner as possible. While discussing these issues could take us in a whole new direction, I am going to continue to focus on data quality here.

It is quite clear that crowd sourcing of map data can produce real benefits. What is less clear is how to evaluate whether cartographic data (crowd sourced or not) is fit for a specific use. What’s worse, I am not sure that the professionals who should be concerned with the concept of fitness of use have yet to provide valuable metrics to measure the efficacy of map data for either general or specific applications. Maybe that is as it should be? I doubt it.

All of us can remember looking for the “best” map dataset we could find for a specific research topic. In many cases, we had to settle for less than the best, because it was unavailable to us for a variety of reasons. So, we made a concession here and there and used the map to create a geographic base for the story we wanted to tell. I suspect more concern went into analyzing the data than evaluating the map. In a way, the process is not unlike those home repairs we all attempt. Sometimes do-it yourself jobs look pretty good, but we know where we screwed up and where we had to take short cuts to get the danged thing done. I think maps and mapping likely exhibit the same types of “hidden” mistakes.

Often, even when we had access to the “official” data, it was none too good. Yet we made do. Google Map Maker and OpenStreetMap are more recent examples of “making do”.

It is likely that local people can provide the best data for local areas. If you can assemble knowledgeable local people over a number of local areas, then you can extend your map coverage to considerable spatial extents. However, map databases that spatially autocorrelate with population density leave a lot of the world unmapped. In addition, there are clearly some governments who think maps are a strategic advantage and won’t allow crowd-sourcing. In addition, there are great swaths of the planet where the internet isn’t part of the daily routine and all the great, new mapping tools will not make much of a difference in map data availability or quality.

In other words, there are spatial and technical limits to the availability of crowdsourced data. I think that there are also “sensor” limits to crowdsourced data. You know, the things that are mapable but cannot be seen (e.g. borders), or things that can be badged/signed but do not bear the name everywhere (e.g. rivers, lakes, mountain ranges) and perhaps not near where the “crowdsourcers” are working.

So, if crowdsourced data gathering and do-it-yourself mapping are replacing the professionals, who is going to provide the gatekeeper functions they once provided? Maybe that is what companies like CloudMade are going to do? (By the way, Marc Prioleau, formerly of deCarta has just become CEO of CloudMade.) On the other hand, maybe there just won’t be gatekeepers.

When we started out on this gatekeepers journey, we started by discussing the concept of authoritativeness. Crowdsourcing and the “it’s about me” generation may turn this on its head. While I may feel comfortable with a gazetteer that includes the commonly accepted name for a geographic location, it is possible that other people may feel that another name is more appropriate, or simply prefer another name. Gazetteers would simply become a repository for all names that might be associated (either correctly or incorrectly) with a location.

Similarly, streets could have alternative names (even designer names) and all of those could be stored and used by “whomever” for “whatever”. Countries can even be given different names of different days. Oh, wait, that won’t work will it? Isn’t there some notion about maps matching reality, not the one we want, but the one we observe?

Whew, so that means there is still the need for Gate Keepers. Unfortunately, nobody seems to be willing to step-up and play the role. OK neogeographers – that’s the challenge you face. Best of luck. (Well, I was going to tell them how you could achieve this goal at the Where2.0 conference, but my proposal to speak there was turned down. Drat – now I do not have a good reason to go – and it is a fun meeting!)

Of course, I will be giving an invited presentation at GPS Wireless at the end of this month, if you are interested in hearing my dulcet tones.

Now, on to a couple of minor notes

When I was researching the Google Map Maker stuff I ran across a great map of Sri Lanka. – you can find it here.

Speaking of interesting maps – check this one out.

Next, I was contacted by a fellow named Rob Sequin, who lives on Cape Cod. Seems he has a number of interesting domain names related to LBS that are for sale – hmmm. In fact he has LocationBasedServices.com. If you are looking for a good name for a location based business, you might want to take a look at his inventory.

Of course, if you like Rob’s names, you might like my GeoTagmyBiz.com or LocalizemyBiz.com – but I’m holding out for a spot on the Internet Millions Infomerical.

Bookmark and Share

Posted in CloudMade, crowdsourced map data, Data Sources, geographical gazetteers, Google Map Maker, map updating, Mapping, Mike Dobson, User Generated Content


(comments are closed).