There are plenty of route-planning services showing you how to get places quickly. But what about getting to those places the most happily?
We’ve all taken a detour because the path is pleasant and scenic, even if it takes longer. But Google Maps and the like aren’t set up for that. They’re solely about speed and efficiency.
Recent research led by Yahoo Labs shows how a planner-for-happiness might work. Using crowdsourced impressions of streets, Flickr data, and survey responses, it looks for a balance between “people’s emotional perceptions of urban spaces” and getting them to a destination in a reasonable amount of time.
“To date, there has not been any work that considers people’s emotional perceptions of urban spaces when recommending routes to them,” the paper, which is titled “The Shortest Path to Happiness,” notes. There are route-planners that deliberately take in tourist sites or amenities, or even aim to get people lost so they have a higher chance of serendipitous experiences. But they don’t measure responses to surroundings, the researchers point out.
Led by Daniele Quercia, the new work first used a crowdsourcing site called UrbanGems.org. Visitors were shown two London streets side-by-side, then voted on which was more beautiful, quiet and happy. In all, more than 3,000 people responded and the results clearly showed a link between greenery and positive impressions. “We discovered that the amount of greenery in any given scene is associated with all the three attributes, and that cars and fortress-like buildings are associated with sadness,” Quercia says.
The problem with crowdsourcing is that it requires active user participation. To build an instant route planner for multiple places would be hard. So Quercia and his colleagues turned to a more passive source of geo data: Flickr.
Dividing London and Boston into 200 meter squares, they looked at the number of pictures available for each cell, the number of views, favorites and comments, and also the type of language associated with each entry. Then they ran an automated linguistic system to assess tags and comments for positive and negative meaning.
“The system uses geo-tagged Flickr pictures and the associated metadata to build an alternate cartography of a city weighted for human emotions,” Quercia explains in an email. “The idea is that people are more likely to take pictures of historical buildings, distinctive spots and pleasant streets, instead of car-infested main roads.”
To assess whether Flickr is a viable medium for happiness assessment, 30 participants in London and 54 in Boston rated the results versus the crowdsourced responses. While less accurate, the researchers still think the second method is valid.
Quercia is now working on an app that “suggests a path between two locations that is the shortest route that maximizes the emotional gain,” which he hopes to test in European and American cities soon. It can’t be long before we have ways to assess a city’s “psychogeography”–how the environment makes us feel–instead of just ways to navigate it the fastest.