Imaging the City Workshop at CHI

The Imaging the City workshop here at CHI featured a nice set of practices and technologies for representing the urban environment in Human-Computer Interaction. Some of them are strongly related to my current work.

Urban Ritual in Rome: Characterizing the City with High-Resolution Cell Phone Data by Matthew Jull and Carlo Ratti is a follow-up on their work with Real-time rome. Their spatio-temporal analysis show that cell phone activity correlates strongly with the infrastructures and zones of the city (e.g., transportation links, commercial, office, residential zones) and the daily activities (e.g. wake up, commute, work, lunch, dinner, sleep, sports events). These preliminary results suggest that cell phone data can be used to characterize and map urban domains and the cultural signature of its occupants.

The daily activity of people throughout the city is revealed (i.e. wake up, commute, work, lunch, dinner, sleep, sports events etc.)
Personalized City by Continuous Location Logging by Jun Rekimoto and Takashi Miyaki. They propose a city visualization method based on long-term and continuous personal location history generated by WiFi-based location detection methods. Their visualization reveals a person’s “cognitive” map based on the traces left while navigating the urban space. They apply a fisheye-view method to distord the map in order o reflect the person’s location probability density. Therefore, frequently visited area becomes larger on the map. The continuous log of one’s geographical position could be a foundation for various “lifelog” applications, such as memory aid or information organizer (as suggested in Tracing Personal Mobility).

A nice comparison of WiFi in GPS location accuracy in a urban setting

Person’s “cognitive” map based on location history. Fisheye-view to reflect the person’s location probability density. Frequently visited area becomes larger on the map. To be useful in a local search task, it might be more useful to distort the map to reveal the less know areas. I have this experiment using anamorphosis maps in mind for a while now…

How We Watch the City: Popularity and Online Mapsi by Danyel Fisher. He developed Hotmap, a tool that visualizes how people have used Microsoft’s maps.live.com to reveal what parts of the maps they find most compelling. The results show that the vast majority of hits are focused on a fairly small area, following population, suggesting that largely, users currently find use of the tools for looking at natural scenery less compelling then they do city imagery. This work could be extended by show how people “query the city” (e.g. what are the keywords. Something I wanted to find out more about at Urban Mapping). Knowing more about the types of searches and the areas of interests could help the local search community to go beyond the classic, yet irrelevant “closest Starbucks” scenario. Categorizing the different types of granularity in space and time embedded in the queries could help understand how people query space.

The Big Picture: Exploring Cities through Georeferenced Images and RDF Shared Metadata, by Carlo Torniai, Steve Battle and Steve Cayzer. They present a prototype system by on Flickr geotagged images to discover geo-related pictures of a same city taken from different users. The browsing interface, based on different levels of relationship between pictures metadata, allows users to explore cities according to the relations discovered. An extension of the work can be found in Sharing, Discovering and Browsing Geotagged Pictures on the Web”.