Volunteered Geographic Information
Free, Priceless Or Somewhere In Between?
This is the title that has been popping into my head since last summer. I am writing it down because it encompasses in a very general sense the themes that I want to cover in my dissertation, and thus serves me well in trying to guide me while I try to elaborate on them.
I have actually already written some paragraphs elaborating on the themes and ideas that follow, but I want to force myself to touch upon them concisely here.
Volunteered Geographic Information (VGI) is a concept that has not been around for a very long time. Geographic Information has, however: it is what maps are made out of, and what your car navigation device relies on to guide you. Traditionally, Geographic Information is collected, processed and used by professionals, but this no longer holds true: Geographic Information has undergone a process of democratization, both in the usage dimension and in the collection and processing dimension. People are now used to dealing with Geographic Information in different contexts, and have started to pool resources to collectively build repositories of Geographic Information, to facilitate the democratization of the entire ecosystem of Geographic Information.
OpenStreetMap is the most prominent of these efforts, and one in which I have been actively involved since early 2007. Since its conception in 2005, it has grown to a worldwide collaborative effort involving more than 100.000 contributors. In some regions, the maps available from OpenStreetMap are so rich and complete that they are used instead of commercially available map data.
I realize that I need to come up with some examples here, and some numbers that give an indication of how OpenStreetMap has grown, but I am on a train, blissfully disconnected from the internet, so you will just have to bear with me for now. But believe me, it’s getting big fast – at a rate that makes me worried about the validity of any quantitative research results that I might present in the context of this dissertation. But this will have to be dealt with in some future note.
Let us assume for now that OpenStreetMap – there are other VGI efforts around, and they will need to be touched upon as well – is indeed starting to occupy a significant share in the commercial market for Geographic Information. That means the OpenStreetMap data represent a commodity and as such, economic value. As OpenStreetMap data is available at no cost, this value is not quantified in the marketplace, however. This poses intriguing questions:
What is this freely available OpenStreetMap data actually worth?
How do you even begin to measure the value of something that is not subject to the usual economic market mechanisms?
When dealing with value, I believe I cannot omit the concept of quality, especially in this context. Any VGI effort relies on volunteers collecting data in their spare time. While some regions have very active communities, getting together to discuss progress and plan improvements to the map, checking and correcting each other’s contributions, other regions rely on single, isolated individuals contributing to the map – or worse: no-one contributing at all. The resulting picture is one of spotty coverage: very densely mapped regions exist side by side with tersely covered regions. More questions arise!
Is it possible to define the quality of volunteered geographic information in any satisfactory way?
More generally: how do quality and value relate when dealing with geographic information?
I think I cannot proceed from here without looking at real world situations. Economic value is defined in the marketplace where supply and demand meet, and thus cannot be studied without some understanding of how and where this demand arises.
There clearly is a demand for VGI, but where does it originate?
Why would people want to use information that comes with no guarantees of completeness or even factual correctness, and that does not have a consistent quality?
I will need to get to the bottom of this. Apparently it is ‘good enough’ for some! If I’m not careful I will be entering into the domain of psychology. I think I need to stop soon, or I will have covered all domains of modern science and will have defined ample questions to last me three dissertations. But let me just finish this train of thought, and by then I will have arrived in Berlin – one of the best covered cities in OpenStreetMap, by the way; you can even get a detailed map of the zoo!
What drives the decision on the demand side to use volunteered geographic information instead of commercial offerings that do come with a quality label?
I can think of a number of reasons. Firstly, there is a growing number of application domains that do not require extensive, nationwide coverage. The growing domain of location based services are often only relevant in metropolitan areas; consider for example pedestrian and bicycle routing, social networking applications, tourist guide services or restaurant / bar recommendation applications. Even many applications in professional domains operate only within a designated metropolitan area: local police, fire brigades and other public safety professionals operate only within their metro area.
Interestingly, supply and demand sync up really nicely here: in areas where there is likely to be a great demand for high quality – whatever that may mean – geographic information, there is also likely to be a large number of contributors to volunteered geographic information repositories. (This reminds me of my master’s thesis that dealt with the quality of public transportation in rural areas. There was a similar process at play: because of the limited and geographically thinly spread demand, the costs of maintaining a reasonable quality of service had become so high that cuts in service quality had become unavoidable, lowering the demand even further. Both Dutch and German regional governments were struggling to counter this downward spiral, and I did a comparative study on the results of those efforts.)
Secondly, because there is very little restrictions and limitations in terms of how and where you can use the data. Commercial data usage licenses are more often than not restricted to a certain type of application, device or to a limited number of users or devices, and the data can only be used as-is. OpenStreetMap data can be used in almost every context imaginable, and you are free to modify and adapt the data to suit your needs.
Lastly, of course, because it’s free.
I have mixed feelings about this post. It feels unfocused, but I guess that is to be expected. More importantly, I don’t feel comfortable in the domain of economics. Sure, I did my two years of high school accounting and economics, but it did not quite take. It does not particularly interest me, but I feel I need to deal with it anyway. Intuitively, I am drawn to the question of defining and measuring quality. I want to think about how to do that, write tools to analyze OSM data – that part I am really passionate about. It seems like a good moment to talk to Henk and maybe some other people I know that could help and advise me at this junction.