AI helps unravel the secrets of Europe’s prehistoric ‘green gemstone’ trade – Popular Archeology

University of Seville—A multidisciplinary team of archaeologists and artificial intelligence experts from Spain and Portugal has combined nondestructive archaeological measurement techniques, machine learning, and explainable artificial intelligence (XAI) tools to develop an AI system that can be applied to archaeological research. In this particular case, the aim is to investigate the provenance of an archaeological sample of variscite. Variscite is a mineral with a characteristic green color that was highly valued in prehistoric times and was distributed by extensive exchange networks throughout Western Europe between the 6th and 2nd millennium BC. It was used to make general jewelry such as necklaces, bracelets, and rings.

This group of researchers has been working together for years to find out where varisite came from among various archaeological sites on the Iberian Peninsula. To do this, they compare current geological samples of varisite with samples found in archaeological excavations. They analyze minerals, record their elements, and compare the small chemical changes they exhibit. Based on similarities, you can determine where it was extracted from.

This study* archeology journalled by the University of Lisbon, with participation from the Mira y Fontanals Institute of Humanities (IMF-CSIC), the University of Seville, the University of Alcalá, and CIPAG (Spanish acronym for “Prehistoric and Archaeological Research Group of Garraf Ordal”).

Unique geochemical footprint of each mine

The innovation of this study lies in the use of AI to analyze the chemical composition results. “Our model learns to recognize each mine’s unique geochemical footprint, allowing us to determine where prehistoric beads came from, even thousands of years after they were produced.” Daniel Sanchez Gomeza researcher at the University of Lisbon and lead author of the study. Thanks to this pioneering approach, we were able to predict the geological origin of archaeological objects made of varisite with 95% accuracy.

In this way, the team built the most extensive compositional database ever created, including more than 1,800 geological samples and 571 archaeological accounts analyzed using portable X-ray fluorescence.

To process the data, they used a random forest algorithm. this is, machine learningwhich allowed us to achieve unprecedented accuracy. To explain further, Ferran BorelIMF-CSIC archaeologist: “What’s really remarkable about this project is that the information was uploaded to Zenodo.” [an open repository developed under the European OpenAIRE program and operated by CERN]Allow other researchers to use the data to make their own interpretations. We are working in collaboration with open science. ” Ferran Borel He is leading an excavation project for a prehistoric varicite mine in Gaba (Spain), from which some of the samples studied are taken.

As a result, it has become possible to reinterpret prehistoric trade routes. Now, researchers explain that it can be seen that the Gaba Mine (Barcelona, ​​Spain) and the Ariste Mine (Zamora, Spain) were major production and distribution centers. The traditionally cited source of Encinasola (Huelva, Spain) would have been less important. And the material found in Brittany (France) likely came from the north of the Iberian Peninsula, suggesting an overland route across the Pyrenees rather than the previously proposed sea route.

“We used explainable artificial intelligence techniques, which allow us to explain in a clear and understandable way how AI models, especially the most complex models, make decisions. In the case of our study, this means not only making accurate predictions, but also showing which chemical elements are decisive in each classification, bringing transparency and rigor to archaeological interpretations.” Carlos OdriozolaProfessor at the University of Seville and PI of the project. This methodological framework, called VORTEX (X-ray-based Variscite Origin Recognition Technology), opens new possibilities for the study of the provenance of other archaeological materials, such as amber, and constitutes a milestone in the application of artificial intelligence to cultural heritage.

Well, according to Manuel Ed Venagewe, the co-authors of this article, have to address the following questions: What are the reasons for the phenomenon of green stone expansion? How did this expansion into Western Europe occur over time? Where did it start? Interpretations are numerous, but always with the common goal of knowing more about the past. “This is not just a matter of green beads; it is about using artificial intelligence to tell prehistoric stories to humans,” he concludes. Sanchez-Gomez.

____________________________________________

Article source: University of Seville News release.