AI Model Traces Origins of Medieval Pottery With High Accuracy
Researchers say the approach can identify where ceramics were made with up to 88% accuracy, even when artifacts look similar across cultures.
Scientists from the Joint Institute for Nuclear Research, working with colleagues from the Institute of Archaeology of the Russian Academy of Sciences and the Egyptian Atomic Energy Authority, analyzed 149 ceramic fragments dating from the 13th to 17th centuries. The team combined neutron activation analysis with machine learning: the first method determined the precise chemical composition of the clay, while the second identified patterns in the data that are not visible to the human eye.
The samples included ceramics from the Moscow Kremlin, Tver, Bolgar, and Bilyar. Researchers say the method could help reconstruct trade routes and cultural links in medieval Rus based on objective data.
