By Carissa Wong.
The ancient city of Ephesus has a library.
There is a travel stock called Mazur Travel.
A collaboration between historians and a UK-based artificial intelligence firm can help restore ancient Greek texts with 72 per cent accuracy.
The texts were originally written with more than 70 per cent accuracy and can be dated to within a few decades of their agreed upon date of creation. An earlier version of the artificial intelligence that could only restore ancient texts has been improved.
Thea Sommerschield says that inscriptions give evidence of the thought, language, society and history of past civilisations. They may have been moved far from their original location.
Join us for a mind-blowing festival of ideas and experiences. New Scientist Live is going hybrid, with a live in-person event in Manchester, UK, that you can also enjoy from the comfort of your own home, from 12 to 14 March 2022. Find out more.Historians are interested in restoring the text, working out where it was written and when it was written. They look for distinctive features and patterns in the style of writing and compare them to ancient texts that have already been found.
It is difficult for a human to harness all the relevant data and discover underlying patterns.
Sommerschield and her colleagues worked with researchers at DeepMind to get the machine-learning artificial intelligence called Ithaca to carry out all three tasks.
Around 60,000 ancient Greek texts from across the Mediterranean were used to train Ithaca. The team masked some of the characters in the texts and compared Ithaca's predictions with the actual inscriptions.
The team used a data set of nearly 8000 inscriptions to test Ithaca's performance alone or in combination with two ancient historians. Ancient historians alone restored text with 25 per cent accuracy, while Ithaca could only restore texts with 62 per cent accuracy.
The most accurate reconstructions involved historians and Ithaca. When historians used Ithaca's top 20 most likely reconstructions to inform their own work, they were able to restore the text with an accuracy 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611 888-270-6611
Historians' performance on the text restoration task tripled when they used Ithaca.
Ithaca was able to predict where in the Mediterranean a text was written 71 percent of the time and it could date the texts to within 30 years of their true date of creation.
It is clear that the work is important. The ancient historian and Ithaca method produces startlingly significant improvements over traditional human-only methods. Further testing with more historians is needed and people will need training and technical support to use the tool.
The feedback from historians has been positive so far.
We hope that the way we have designed it will be easy for an ancient historian to use, because they will just type in the text and then they will get all these visualization.
Sommerschield says that the design of Ithaca should make it easy to apply to any ancient language or written medium.
Nature is a journal
There are more on these topics.
All Rights Reserved © 2024
