Monthly Archives: January 2012

Reflecting on last year’s work and events, I could noticed a gradual shift in terminology used in our publications: We’ve basically gone from using “VGI” (for Volunteered Geographic Information) to using “UGGC” (for User-Generated Geographic Content) instead. So what, you might think? Yet another acronym, duh. I mean, with the advent of social networking and social media platforms, we have witnessed (and still are witnessing) the hype of various terms. One of the latest is “geosocial” (yes, I have jumped on the bandwagon here…). Each of these terms has certain issues. UGC is the broadest term of them all and encompasses everything. Does it have anything to say at all? VGI is not widely known outside academic circles, but has created a comparable hype there. Since its inception by Mike Goodchild, this term has been used in a wide array of research. I have to admit that I’ve always had my issues it, because it seemed too narrow to capture all the aspects of user-generated content with a geographic component. While the term certainly describes well the effort of communities like OpenStreetMap, it does not seem to work well with other media, e.g. Tweets. Are public Tweets volunteered information? If they contain a place name, are they “geographic”? These reservations did not stop me from using the term VGI in the context of my research, which at the moment primarily deals with Tweets and Flickr photographs as input data. At the time, It did not seem wise to introduce yet another term. More than a year later, we were using UGGC more and more. This prompted me to spend some more time on rethinking whether this was actually a good move, and what its implications might be. There are two aspects that led to our change in terminology. One is how citizens share their knowledge, the other is the old question “What’s geographic?”:

The new technologies summarized under the vague umbrella-term “Web 2.0” enable every citizen with basic literacy and access to the internet to contribute to and consume a common body of knowledge. This body of knowledge has many forms, from Wikipedia to OpenStreetMap, to the ubiquitous rating and review functions of travel-related search portals, movie databases, recipes sites, auctions, etc.. The number of initiatives gathering, curating and disseminating explicitly geographic information has exploded with the free availability of online maps (in particular Google maps in 2005, compare Elwood, Goodchild, and Sui 2011), which can be embedded in mash-ups with other content. The central point here is that the quantity and availability of information has increased dramatically, but is this information truly volunteered? For example, what about the content that is published in social networks and micro-blogging services? Sure, this is shared, but is it truly volunteered? According to Merriam-Webster online, a volunteer is “a person who voluntarily undertakes or expresses a willingness to undertake a service”. So, volunteerism must have a specific goal or a purpose and is carried out in the framework of a service (or portal). Crowd sourcing forest fire information by offering a portal where citizens can send their reports (in whatever form) relies on (explicitly) volunteered information. A portal that harvests such public reports from various sources without the explicit consent of the authors is using “only” shared information (implicitly volunteered).

Regarding the issue of geography: With many social media platforms adding the ability to geographically reference any published content information (e.g. Twitter GeoTagging, Facebook Places), and the wide-spread use of GPS sensors in portable devices like digital cameras and smartphones, we can expect to encounter much more georeferenced Web 2.0 content. This leads to the question, whether this turns the content actually into geographic information. A common assertion in the geographic information industry is that 80% of all information is related to geographic features. This assertion stands on shaky foundations (compare Hahmann, Burghardt and Weber 2011), and it leaves open the question whether a geographic reference is enough to make information geographic. It seems to make sense to distinguish between geographically enriched information (implicitly geographic), and truly geographic information that is about a place (explicitly geographic).

So, to recap, I suggest to consider whether the information was truly volunteered or only shared with the public, and whether the geographic aspect is only a reference or core the theme of the information. The combination of these two dimensions gives us a matrix of four types of information:

  1. Explicitly geographic and explicitly volunteered: This is “True” volunteered geographic information (VGI) in the strictest sense. Examples include Open Street Map.
  2. Explicitly geographic but implicitly volunteered: User-generated geographic content (UGGC). Examples would include any public Tweet referring to the properties of an indentifiable place.
  3. Implicitly geographic but explicitly volunteered: Volunteered (geo)spatial information (VSI). Examples would include Wikipedia articles about non-geographic topics, which contain place names (e.g. articles about a person, mentioning place of birth, etc.).
  4. Implicitly geographic and implicitly volunteered: User-generated (geo)spatial content (UGSC) such as a Tweet simply mentioning a place in the context of another (non-geographic) topic.

And of course the rest (or umbrella term?) of UGC for anything not captured by the above.

Following this, it seemed obvious that we had gone from VGI to UGGC, when we collected Tweets about forest fires: A forest fire can be viewed as the property of a place (forest is burning or has burnt), and unless sent (volunteered) to a specific portal, it is not explicitly volunteered (although the authors certainly expect it to be read).

We have discussed this a lot, but there are still many open questions or borderline cases. Does it make sense to use “geospatial” as distinctive from “geographic”? For me, yes, with “geospatial” enabling georeferencing and “geographic” describing properties of a place, but there has been an unfortunate blending (some would say confounding) of these terms in the literature. Further, does it make sense to distinguish between content and information? Finally, “User-generated” is problematic because it implies a hierarchy between user and system – try imagining Web 2.0 platforms without users – what’s left?

The gentle reader is encouraged to voice his/her thoughts!