AI Translation of Hittite Texts, in Bulk

Discuss the world of the Greeks, Romans, Babylonians, and Egyptians.
Post Reply
User avatar
billd89
Posts: 844
Joined: Thu Jun 25, 2020 6:27 pm
Location: New England, USA

AI Translation of Hittite Texts, in Bulk

Post by billd89 »

https://arkeonews.net/turkish-researche ... e-tablets/

500 Hittite cuneiform tablets were translated at the start of the project by photographing them in high resolution and scanning them with 3D technology. According to the results of the testing, the AI’s success rate was 75.66%.

User avatar
billd89
Posts: 844
Joined: Thu Jun 25, 2020 6:27 pm
Location: New England, USA

Re: AI The Fragmentarium

Post by billd89 »

https://phys.org/news/2023-02-ai-texts- ... dable.html

The Babylonians wrote in cuneiform characters on clay tablets, which have survived in the form of countless fragments. Over centuries, scholars transferred the characters imprinted on the pieces of clay onto paper. Then they would painstakingly compare their transcripts and—in the best case—recognize which fragments belong together and fill in the gaps. The texts were written in the languages Sumerian and Akkadian, which have complicated writing systems. This was a Sisyphean task, one that the experts in the Electronic Babylonian Literature project can scarcely imagine today.

Digitization of all surviving cuneiform tablets
Enrique Jiménez, Professor of Ancient Near Eastern Literatures at LMU's Institute of Assyriology, and his team have been working on the digitization of all surviving cuneiform tablets since 2018. In that time, the project has processed as many as 22,000 text fragments.

"It's a tool that didn't exist before, a huge database of fragments. We believe it can play a vital role in reconstructing Babylonian literature, allowing us to make much faster progress." Aptly named the Fragmentarium, it is designed to piece together fragments of text using systematic, automated methods. The designers expect that the program will also be able to identify and transcribe photos of cuneiform scripts in the future. To date, thousands of additional cuneiform fragments have been photographed in collaboration with the British Museum in London and the Iraq Museum in Baghdad.

An algorithm discovers new texts and matches up fragments
The team is training an algorithm to piece together fragments that have yet to be situated in their proper context. Already, the algorithm has newly identified hundreds of manuscripts and many textual connections. In November 2022, for example, the software recognized a fragment that belongs to the most recent tablet of the Gilgamesh epic, which dates from the year 130 BC—making it thousands of years younger than the earliest known version of the Epic. It is very interesting, remarks Jiménez, that people were still copying Gilgamesh at this late period.

In February 2023, the LMU researcher will publish the Fragmentarium. For the first time, he will also release a digital version of the Epic of Gilgamesh. The new edition will be the first to contain all known transcriptions of cuneiform fragments to date.

Since the project started, around 200 scholars worldwide have had access to the online platform for their research projects. Now it is to be made available to the public as well. "Everybody will be able to play around with the Fragmentarium. There are thousands of fragments that have not yet been identified," says Jiménez.

Also, this
https://www.ancient-origins.net/news-hi ... ai-0017893
Post Reply