Abstract

This paper aims to compare network construction techniques and evaluate which one achieves the best representation of a local provisioning system that connects rural settlements with military castella in the Dutch part of the Roman limes. Using an existing site dataset and a dataset of least-cost path reconstructions between all sites, a number of network construction techniques are described and applied, including maximum distance networks, proximal point networks, a Delaunay triangulation, a Gabriel graph and efficiency networks. They are evaluated using the network metric of average path length and ‘local’ average path length to reach the castella, along with other indicators. Ultimately the Gabriel graph and proximal point networks with a high number of neighbours proved to be the best representation through a good performance on the evaluated indicators as well as the presence of a number of downsides in the other networks, with the Gabriel graph being slightly better due to a smaller number of links needed. This study thus shows that the choice for a network construction technique in archaeological case studies is important and presents a possible strategy to approach such a problem.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call