Data Mining Social Media For Science

What do whale sharks, NASA’s Hubble Telescope, and your vacation videos have in common? Plenty, according to a report by ZDNet’s Jo Best. The author describes in incredible detail the ability to track one of the most elusive (for its massive size) widely-known ocean dwellers, the whale shark.

Whale sharks got a recent “awww, isn’t that cute” public-eye boost thanks to the portrayal of Dory’s childhood friend in the summer film, Finding Dory. The largest living fish and one of three filter feeding sharks that live on plankton, krill, and larvae, the whale shark reaches a length of forty feet or more, and tends to weigh in at the 41,000-pound mark. Unfortunately, their status is listed as “vulnerable” by conservationists, making sightings of this filter-feeding plankton eater all the more important.

As Best describes, whale sharks can be distinguished from one another by their unique spot patterns around their pectoral fins, practically as unique as a human fingerprint. But with so few sightings compared to other heavily researched animals, scientists are turning to an unlikely source for more records of their study: your vacation videos.

By looking through YouTube and other sources for footage of whale sharks, researchers at Wild Me’s WhaleShark.org project can feed the images through their software based on NASA-assisted algorithms. They then compare the images to their existing database of likely sharky suspects. Best describes efforts that are already underway to use this same technology to identify dolphins by their dorsal fin shapes, sizes, and notches, which is good news for science since there are arguably far more tourist and amateur images out there of dolphins than of whale sharks.

Data mining through social media by researchers is certainly nothing new, but so far the efforts have been decidedly darker. Law enforcement has used YouTube footage and cloud-based surveillance videos to place suspects at the scene, for example. A new app called TraffickCam is asking travelers to snap a photo of their hotel rooms each time they travel; the photos are uploaded to a database of hotel rooms so that mined images of sex trafficking victims can pinpoint the location by comparing the room against the images in the database. TraffickCam has already been downloaded more than 56,000 times by smartphone users, just since its launch in late June of this year.