How Remote is Your Sensed Data?

Researchers Fabien Girardin and Josep Blat of Pompeu Fabra University in Barcelona are doing some interesting geospatial analysis using what is essentially "found" geodata.

In Tracing the Visitor's Eye, Girardin and Blat are using the geotagging of photo locations by tourists in the photo-sharing website flickr to populate spatial analysis models and examine how tourists interact with several major cities.

Flickr allows users to "place" their photos on maps, creating simple geodata that can be viewed and used by other applications (assuming a user has set permissions to "public," as many do). Blat and Girardin are working on several research goals with these data:

• Evaluating the potential of using people-generated geotagged information to contribute urban understanding.
• Studying how people explicitly position and disclose spatio-temporal information in order to understand their use and need of quality of location information in a urban space.

That is, can we use tourist-generated geodata to study cities? And, if so, what are the effects of variation in their dependability as geo-locators?

It is the case that flickr users vary widely in their attention to exactness when mapping their photos. The researchers are working on whether or not that variation in data quality can be overcome in analysis.

If so, there are some interesting possible applications. Urban planners might be able to determine which are the most important view-sheds in an area and which need the most protection. State and national parks managers can study what is most interesting to visitors.

Aside from heat-maps of raw tourist interest, the project includes traces based on temporal data included in the photo metadata. These allow researchers to track where and when visitors are going. That led to an interesting insight into the different approaches taken by foreign and domestic visitors to Italy, as reported in a one-page PDF summary prepared by the researchers:

For instance ... Americans follow a specific graph constituted by the nodes of Florence, Siena, Pisa, Genova and Perugia. By contrast, Italians ... are more adventurous in their exploration of the area (including reporting on visits of the Island of Elba).

If nothing else, this research illustrates the expansion of new and public sources of data and information possible with the spread of open-source tools and applications on-line. And it re-enforces the value that professional GIS folks can bring to the analysis and use of that data.