Personalizing Tourist Maps

Floraine Grable presented her work on automating tourists maps at last week’s SIGGRAPH conference.  While street level imagery as Google’s Street View and Microsoft’s Pictometry data are popular, Grable found that providing too much information isn’t helpful for tourist maps, stating, “When the maps show every building, every street, it’s very difficult to find specific sites.“  Grable, who is based at UC Berkeley, worked with ETH Zurich to develop software that would automate which buildings to include on a map depending on the focus of the tourist map.  In San Francisco all buildings were assigned a score based on their relevance as a tourist, shopping or dining site.  All buildings scoring above a threshold would then be automatically added to the map based on the category selected.

Read more:

• Personalised maps show the view from the street – New Scientist Tech
• Automatic Generation of Tourist Maps – Research paper by Floraine Grabler, Maneesh Agrawala, Robert W. Sumner, Mark Pauly

On the left is a map generated by Microsoft Live showing all locations. The map on the right highlights only the tourist sites.