A collaboration between the University of Chicago's Oriental Institute (OI) and the university's Department of Computer Science could enable automated transcription of ancient tablets of Persia's Achaemenid Empire.
The DeepScribe project will use a training set of more than 6,000 images from the Persepolis Fortification Archive to build a model that can "read" the cuneiform characters and potentially be adapted to other studies of ancient writing.
A database of more than 100,000 identified individual signs was used to train a machine learning model, which, when tested on tablets not included in the data set, successfully deciphered cuneiform signs with about 80% accuracy.
Said OI's Susanne Paulus, "If the computer could just translate or identify the highly repetitive parts and leave it to an expert to fill in the difficult place names or verbs or things that need some interpretation, that gets a lot of the work done."
From University of Chicago
View Full Article
Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA