Evaluation of Prehistoric Pottery by Means of Mathematical Models

dr. V. Salač, CSc. 2002 – 2005

A major problem of contemporary archaeology is to manage the rapidly growing number of findings in order to foster further development of this scientific area. Archaeology is able to create large database systems of artifacts, nevertheless the analysis and interpretation of data reaches the limits of individuals, workplaces, even the research area itself. This project struggles to create automatic knowledge acquisition methods by means of mathematical modelling which would evaluate certain features of prehistoric pottery (age, quantity, degree of fragmentation). Artificial neural networks and genetic algorithms will be used to create these methods, probably for the first time within this context. The project outputs aim to a new approach to established archaeological theories concerning not only prehistoric pottery but also the origin of fillings of archaeological features, layers, etc. The project oversteps mathematical applications in archaeology done until now.