Friday, April 22, 2011

‘Citizen scientists’ help search for tomb of Genghis Khan via photos of Mongolia

Albert Lin is hunting for Genghis Khan.

Legend has it that Khan, the ruthless conqueror who was the first emperor of the Mongol Empire, was buried in an unmarked tomb in northern Mongolia about 800 years ago.

But finding said tomb is a task that has eluded scientists for years. Mongolia encompasses more than 600,000 square miles of largely uncharted, rural territory, which makes Lin’s mission an extremely challenging one.

Luckily, the explorer and research scientist at the University of California at San Diego has more than 7,000 people around the world helping with his mission, called the Valley of the Khans Project. The idea is to find the tombs of Genghis Khan and his descendants, and other ancient Mongolian artifacts.

Lin’s army of helpers are amateurs, working from the comfort of their home computers.

Through a Web site called Field Expedition Mongolia, which Lin and his colleagues developed jointly with National Geographic, volunteers are helping sift through 85,000 high-resolution satellite images of Mongolia.

Every time volunteers log in to the site, they are shown some of these images. An online tutorial instructs them on how to look for particular objects and tag them as “roads,” “rivers,” “modern structures” or “ancient structures.” They can zoom in and out and scroll in all directions.

They are also told to simply tag places as “other” if they see something peculiar. This is the sort of vague judgment that humans can perform but that computers cannot, Lin said.

“What a computer can’t do is look for ‘weird things,’ but when you ask a human brain, you don’t have to tell it what ‘weird’ is; we know,” Lin said.

Those weird things could be important archaeological finds, he said.

Last summer, Lin and his colleagues were in Mongolia inspecting the places that had been tagged by the online volunteers. Anytime there was a cluster of tags marked as “ancient structure” or “other,” they would note the latitudinal and longitudinal coordinates, grab their GPS devices and scope it out.

“We’d literally jump on horses or get in a helicopter and go check it out,” said Lin. “Every tag was weighted on how many other people tagged the same thing.”

Projects like this one mark a new twist in “citizen science,” where new technology, when used effectively by large groups of people, can help speed up scientific developments, reduce costs and increase efficiency.

Sometimes online volunteers led the explorers to disappointing finds, such as a herd of sheep on a satellite photo that looked like an ancient structure. But there were also some remarkable ones, such as the discovery of 3,000-year-old Bronze Age tombs, remnants of large cities and ancient monoliths hidden in the region’s vast, grassy steppe.

“These are hard to find on horseback, but from space and in the images, you can make out these shapes,” Lin said.

Making it fun

Though professional scientists have collaborated with amateurs for decades, social networking and the Internet are making it more fruitful than ever.

“We found that we could make something that was engaging enough to inspire people to participate without having to pay them,” says Lin. “This is the part of citizen science that is most interesting to me: How can we motivate people to dedicate their time?”

How? By making it fun, Lin said.

Lin began thinking about creating an online expedition that tied into his real one about five years ago, when’s Mechanical Turk made its debut.

An online crowdsourcing marketplace, Mechanical Turk allows requesters with small tasks to pay people for their time. Anyone with an Amazon account can participate, and the tasks are usually quite simple, such as “pick out the images with tattoos from this set,” or “verify the existence of these business Web sites.” Some tasks pay just pennies, per task or verification, while others pay more.

Lin believed that he could get more traction by creating a site that offered a fun experience rather than a paid one. “People are so excited to learn about Mongolian archaeology,” he said. “They start to learn stuff about what they’re doing and feel more connected to what’s going on in that part of the world.”

Every volunteer who logs on to the Valley of the Khans project site, developed with a design company called Digitaria, gets to feel like an explorer, digging through images and playing what feels like a game but performing work that has much more significant ramifications.

“It connects you more on a personal level than going to a museum,” said Allison Shefcyk, a 24-year-old in Connecticut who tagged more than 50,000 images from her home computer. “I ended up picking up some books on Genghis Khan and Mongol culture, and even though I never set foot there, it all provided a deeply moving experience."

Biomedicine and galaxies

The Khan expedition is not the only research project engaging people on their home computers with gaming strategies.

Another is EteRNA, created by researchers at Carnegie Mellon and Stanford universities. It allows players to tinker with nucleotide bases and come up with synthetic RNA designs.

The creators of the site hope that by generating a large assortment of designs, they can speed up discoveries in biomedicine. Every week, the most promising designs are actually synthesized by scientists in a lab at Stanford.

An older project, Galaxy Zoo, allows volunteers to help classify images of galaxies taken by a robotic telescope in a project called the Sloan Digital Sky Survey.

There were thousands of images to classify, so researchers decided to create a crowdsourcing tool that uses the public’s help in making classifications.

To participate, volunteers flip through images and answer simple questions, such as whether a certain galaxy looks completely round, partially round or cigar-shaped.

A year after its July 2007 launch, Galaxy Zoo had more than 50 million classifications from more than 150,000 people. The project so far has generated more than 20 academic papers by researchers around the world in astrophysics and astronomy journals.

And birds, too

Another project, run by the U.S. Geological Survey, is using crowdsourcing and technology to digitize a project called the Bird Phenology Program.

The project was started by Wells Cooke, an American ornithologist who wanted to gather information on bird migration. Starting in 1881, amateur bird-watchers mailed in thousands of index cards detailing information on birds they had seen, first to Cooke and then to the American Ornithologists’ Union. The federal government maintained the program in its final years, but participation declined and it was closed in 1970.

But the cards, and the wealth of information they contain, remain in file cabinets. The U.S. Geological Survey is scanning the cards, and volunteers can log on and enter the information into a central database.

It’s secretarial work, not rocket science, said Jake Weltzin, executive director of the National Phenology Network, the USGS-funded organization that is coordinating the program, but “people really like to contribute to the scientific process and really feel they are contributing to a bigger project,” Weltzin said.

“These are all volunteers; one person has digitized 20,000 cards,” Weltzin said. “We couldn't afford to pay her.”

Eventually, researchers will be able to use the data to study population changes, and perhaps better understand the effect of climate change on birds and their habitats.

Shefcyk, the enthusiastic volunteer in the Valley of the Khans project, said that participating was particularly meaningful to her.

As a child, Shefcyk had trouble fitting in with other kids and was eventually identified as having an autism spectrum disorder. But she spent her time devouring books on archaeology and reading nonstop about dinosaurs, Egyptian pyramids and Mayan ruins.

“I didn’t have many friends, and archaeology was the whole world to me,” she said. “I would always start conversations and ask things like ‘Who’s your favorite Mayan king?’ ”

Flash forward 15 years and Shefcyk is still fascinated by archaeology and is a frequent visitor to the National Geographic Web site. When she found out that ordinary people could help Lin with his project, she was intrigued. So every day last summer, while Lin’s team was in Mongolia, she would tag images and look out for blog posts from the scientists. She was enthralled to read their chronicles and get quick feedback on the sites they visited. The explorers also offered tips to help volunteers tag objects more accurately.

“It’s one of those things where you’re adding your piece, and it’s about knowing that you’re something that’s much bigger than yourself, no matter how small the involvement,” she said.

As for whether any of the tagging has gotten him closer to finding the burial site of Genghis Khan or one of his successors, Lin is coy. “I can’t say yet what we found,” Lin said. “We’re in the midst of compiling the research.”

Bhanoo, Sindya N. 2011. "‘Citizen scientists’ help search for tomb of Genghis Khan via photos of Mongolia". Washington Post. Posted: April 4, 2011. Available online:

No comments: