DATA MINING ANALYSIS OVER A DATABASE OF MATERIALS OF PORTUGUESE PAINTINGS FROM THE 15TH AND 16TH CENTURIES

Abstract

This paper explores laboratory analyses of ground layers of Portuguese wooden painted panels of the 15th and 16th centuries, performed in the Laboratório de Conservação e Restauro José de Figueiredo – Instituto dos Museus e da Conservação. Based on this information, a database of materials was built. A subset of the database was selected to perform a data mining analysis. We used a method based on decision tree learning, Fisher’s exact test and permutation testing and it was possible to find a ground layer technique that distinguishes António Nogueira’s Ferreira do Alentejo retable from other paintings in our database. This constitutes a small contribution to a better understanding of Portuguese painting since it was possible to establish that António Nogueira used coloured ground layers, which was an emerging technique at the time. Another contribution of this work is to show that the methodology we present here can be applied to other case studies based on similar data.