© IE Online Media Services Pvt Ltd
Tags:
Journalism of Courage
The Herculaneum papyri are more than 1,800 papyrus scrolls that were discovered in the 18th century in Italy. They were charred by the explosion of Mount Vesuvius in 79 AD and were long thought to be unreadable. But now, a 21-year-old computer science student won a global competition to read the first text from the scrolls using a machine learning algorithm. .
Luke Farritor, who studies at the University of Nebraska-Lincoln, developed the algorithm that has detected Greek letters on many lines of the rolled-up paper. He used the subtle, small-scale differences in surface textures to train his neural network to highlight the ink. One of the words he deciphered using the machine learning algorithm is πορϕυρας, which means “purple,” according to Nature.
The Vesuvius Challenge gives a series of awards, leading up to a $700,000 main prize for reading four or more passages from a rolled-up scroll. The organisers on Thursday announced that Farritor won the “first letters “prize of $40,000 for reading more than 10 characters in a 4-square-centimetre area of papyrus. Youssef Nader, a second contestant, won $10,000 for coming in second. Nader is a graduate student at the Free University of Berlin.
The scrolls themselves look like wrinkled lumps of coal. Scholars who tried to unroll them after they were discovered in 1752 tried to unroll them but then stopped when they discovered that they were effectively destroying them, reports the New York Times.
“We knew if we could read just one [scroll], then all the other ones would be available with the same method or some augmented method. Some 95 percent of the material from the classical period is lost, so we just don’t have anything, and yet we know it was one of the most important philosophical periods of humanity. It’s an era shrouded in mystery for which we’ve lost most of the material,” said Brent Seales, a computer scientist at the University of Kentucky, to National Geographic.
Casey Handmer, a contestant, won $10,000 in August from the Vesuvius Challenge for the first person to find “substantial, convincing evidence of ink” within the unopened scrolls. While many ancient inks contained metals, the inky used on the Herculaneum scribes was made from charcoal and water, which means that it is barely distinguishable from the carbonised papyrus.
Handmer detailed the discovery of a “crackle pattern” on the scroll that appeared to be ink. Farritor used such crackle patterns to creat the machine-learning algorithm, according to Smithsonian Magazine.
