Ugaritic is an alphabetic cuneiform script known from tablets found in the ancient city of Ugarit (Syria). Of the 30 letters that make up the Ugaritic alphabet, the new software has correctly assigned 29 of them to their Hebrew counterparts.
To accomplish this feat, the system, created by Regina Barzilay and her colleagues, first chooses a nearby language whose meaning is known, in this case Hebrew. The next step is to map the alphabets and compare them for similar frequencies in a systematic way. The app also searches for cognates or words with common roots, such as homme (French) and hombre (Spanish). Despite the automation of the process, the researchers assure that the role of the translator will continue to be important, since it is necessary to take into account the context to resolve ambiguities , something that the program has not yet achieved.
The new software will not only help archaeologists decipher the world’s oldest languages, but it could also expand the number of languages that machine translation systems like Google Translate can handle.