Working in Tandem to Uncover 3D Artefact Distribution in Archaeological Excavations: Mathematical Interpretation through Positional and Relational Methods

Abstract

In recent years, the most advanced pioneering techniques in the computing field have found application in assorted areas. Deep learning approaches, including artificial neural networks (ANNs), have become popular thanks to their ability to draw inferences from intricate and seemingly unconnected datasets. Additionally, 3D clustering techniques manage to associate groups of elements by identifying the specific inherent structures exhibited by such objects based on similarity measures. Generally, the characteristics of archaeological information gathered after extraction operations align with the previously mentioned challenges. Hence, an excavation could be an opportunity to use these prior innovative computing approaches. Our objective is to integrate software techniques to organise recovered artefacts and derive logical conclusions from their spatial location and the correlation between tangible attributes. These results can statistically improve our approach to investigations and provide a mathematical interpretation of archaeological excavations.

Keywords

• 3D density-based clustering techniques
• OPTICS method
• graphic neural networks (GNN) analysis
• 3D artefact spatial distribution detection on an archaeological excavation
• statistical pattern recognition on archaeological depositions
• mathematical interpretation of human occupations