Using a mapping algorithm coupled with citizen reviews of sights and scenery, a team of researchers has developed a way to choose paths through cities based on beauty, quiet and happiness rather than simply theÂ shortest distance between two points.
The project employedÂ Google Street View and Geograph as well as Flickr images and their metadata to build out an initial estimation of probable best paths, then solicited human feedback (to check and enhance the results) from a group of participants on the websiteÂ UrbanGems.
The study,Â published by Cornell Universityâ€™sÂ arXiv, came up with a number of route suggestions in Boston and LondonÂ and contains a number of interesting findings. For starters, the â€˜beautifulâ€™ routes were only slightly longer than the shortest routes, and significantly shorter than typical tourist-oriented directions and guided-tourÂ paths. As the algorithm improves, it is increasingly able to generate paths through new cities via metadata alone, reducing reliance on input from people.