Those pictures of Yellowstone you uploaded to Flickr are doing more than delighting your friends and family — big data scientists are mining the photo website for data that reveals visitation patterns of natural attractions.
When working out how to staff a national park, roadside attraction, or other tourist spot, it helps to know when people tend to visit, how long they stay, and so on. Normally, planners and rangers use on-the-ground methods like checking booking records, turnstile counts, and surveys.
But with hundreds of millions of photos, status updates, check-ins, and other social media being uploaded publicly to the Web, researchers with the Natural Capital Project (operating out of Stanford and the University of Washington) felt there was a vast untapped resource.
The researchers sifted through the millions upon millions of public geotagged photos, resulting in a set of 1.4 million that suited their purposes. And by scraping the (again, public) profiles of the people uploading them, hometowns and other data could be found.
They compared the photo data — mainly when and where these photos were taken — with more traditional surveying methods, and found that it was, indeed, a reliable indicator of when people were visiting. The full results can be read here.
"This is the first study to ground truth the use of social media data to predict these visitation rates," wrote the project's communications manager, Elizabeth Rauer, in a post explaining the research. Now smaller or understaffed locations could get a better idea of expected visits without spending money or manpower on other methods.
It's the kind of "big data" anyone can get behind — at least, anyone who consciously shares their photos publicly, and doesn't mind their data getting scraped for a constructive purpose. This may well help things run smoother at our parks and attractions, provided we can keep them open.
Devin Coldewey is a contributing writer for NBC News Digital. His personal website is coldewey.cc.