Researchers in Israel are hoping to make new discoveries about Jewish history by loading a digital database of manuscripts stretching back 1,000 years into a new transcription tool that uses artificial intelligence (AI).
The Cairo Geniza, the biggest collection of medieval Jewish documents in the world, has been the object of hours of study by academics for more than a century, but not a lot of its more than 400,000 documents have been thoroughly researched.
Although the entire collection has already been digitized and is available online in the form of images, most of its items have not been catalogued, and many are disordered fragments from longer documents.
Photo: Reuters
Only about one-10th have transcriptions.
By training an AI model to read and transcribe the old texts, researchers would be able to access and analyze the whole collection far more quickly, cross referencing names or words and assembling fragments into fuller documents.
“We are constantly trying to improve the abilities of the machine to decipher ancient scripts,” said Daniel Stokl Ben Ezra of the Ecole Pratique des Hautes Etudes in Paris, one of the principal researchers in the MiDRASH transcription project.
The project has already made significant progress and could open up the documents — which are written in Hebrew, Arabic, Aramaic and Yiddish in a wide variety of handwritten scripts — to many researchers, Stokl Ben Ezra said.
Transcriptions from more difficult manuscripts are reviewed by researchers for accuracy, helping to improve the AI training.
“The modern translation possibilities are incredibly advanced now and interlacing all this becomes much more feasible, much more accessible to the normal and not scientific reader,” Stokl Ben Ezra said.
Funded by the European Research Council, the project is based on the National Library of Israel’s digital database of the Cairo Geniza documents and brings together researchers from several universities and other institutes.
One document transcribed by the project is a 16th-century letter in Yiddish from Rachel, a widow from Jerusalem, to her son in Egypt with his reply written in the margins telling of his efforts to survive a plague sweeping through Cairo.
A Geniza is a synagogue’s repository for significant documents that are ultimately intended for ritual burial. The one found in the Ben Ezra synagogue in historic Cairo had a dry atmosphere ideal for the preservation of old paper.
Cairo surpassed Damascus and Baghdad in the Middle Ages as the greatest city of the Middle East, a center of global trade, learning and science and home to a thriving Jewish community, later expanded by refugees fleeing Spain.
The great Jewish philosopher Maimonides, who was physician to the family of Saladin, the Muslim sultan who ousted the crusaders from Jerusalem, worshipped at the Ben Ezra synagogue while living in Cairo.
As dynasties and empires rose and fell, the community quietly went about its daily life, its religious authorities filling the Geniza with the rabbinical arguments, civic records and other detritus of administrative and intellectual business.
The Geniza’s haul of records and papers, including some written by Maimonides himself, was discovered by academics in the late 19th century, but although it has been studied ever since, its enormous size means that huge gaps remain.
“The possibility to reconstruct, to make a kind of Facebook of the Middle Ages, is just before our eyes,” Stokl Ben Ezra said.
ROCKY RELATIONS: The figures on residents come as Chinese tourist numbers drop following Beijing’s warnings to avoid traveling to Japan The number of Chinese residents in Japan has continued to rise, even as ties between the two countries have become increasingly fractious, data released on Friday showed. As of the end of December last year, the number of Chinese residents had increased by 6.5 percent from the previous year to 930,428. Chinese people accounted for 22.6 percent of all foreign residents in Japan, making them by far the largest group, Japanese Ministry of Justice data showed. Beijing has criticized Tokyo in increasingly strident terms since Japanese Prime Minister Sanae Takaichi last year suggested that a military conflict around Taiwan could
A retired US colonel behind a privately financed rocket launch site in the Dominican Republic sees the project as a response to China’s dominance of the space race in Latin America. Florida-based Launch on Demand is slated to begin building a US$600 million facility in a remote region near the border with Haiti late this year. The project is designed to meet surging demand for the heavy-lift rockets needed to put clusters of satellites into orbit. It is also an answer to China’s growing presence in the region, said CEO Burton Catledge, a former commander of the US Air Force’s 45th Operations
Germany is considering Australia’s Ghost Bat robot fighter as it looks to select a combat drone to modernize its air force, German Minister of Defense Boris Pistorius said yesterday. Germany has said it wants to field hundreds of uncrewed fighter jets by 2029, and would make a decision soon as it considers a range of German, European and US projects developing so-called “collaborative combat aircraft.” Australia has said it will integrate the Ghost Bat, jointly developed by Boeing Australia and the Royal Australian Air Force, into its military after a successful weapons test last year. After inspecting the Ghost Bat in Queensland yesterday,
A pro-Iran hacking group claimed to breach FBI Director Kash Patel’s personal e-mail inbox and posted some of the contents online. The e-mails provided by the hacking group include travel details, correspondence with leasing agents in Washington and global entry, and loyalty account numbers. The e-mail address the hackers claim to have compromised has been previously tied to Patel’s personal details, and the leaked e-mails contain photos of Patel and others, in addition to correspondence with family members and colleagues. “The FBI is aware of malicious actors targeting Director Patel’s personal email information,” the agency said in a statement on
RSS