It is impressive for an algorithm to produce even approximate translations of Akkadian into English. But this one still requires significant human intervention:
When they developed the new machine-learning model, they trained two versions – one that translates the Akkadian from representations of the cuneiform signs in Latin script and another that translates from unicode representations of the cuneiform signs. The first version, using Latin transliteration, gave more satisfactory results in this study, achieving a score of 37.47 in the Best Bilingual Evaluation Understudy 4 (BLEU4), which is a test of the level of correspondence between machine and human translation of the same text.Trust me, that first stage of transliterating the cuneiform signs, or even identifying them for unicode, would be a lot of work.
This is amusing:
The program is most effective when translating sentences of 118 or fewer characters. In some of the sentences, the program produced “hallucinations” – output that was syntactically correct in English but not accurate.Back in my postgraduate days, my fellow students and I were well acquainted with Akkadian translation "hallucinations."
I have noted other algorithm projects for cuneiform studies here and here.
Visit PaleoJudaica daily for the latest news on ancient Judaism and the biblical world.
