Application of Artificial Intelligence in Digital Architecture to Identify Traditional Javanese Buildings

Abstract

Traditional buildings have a cultural philosophy and characterize the culture of an area. The occurrence of environmental changes, population growth, and the growth of modern buildings impact traditional buildings. Therefore, preserving those traditional buildings is needed to avoid extinction and make as cultural assets. The research aims to develop an application to help architects quantitatively measure the content of traditional architectural styles in their designs. This study uses the Artificial Intelligence (AI) method to identify buildings' similarities, acquiring traditional building data in roofs and ornaments images as a dataset totaling 650 images of roofs and 7,180 ornaments. Data processing was carried out by making architectural models, training, testing accuracy, and creating application interfaces. The algorithm used to identify similarities between building types was the Convolutional Naural Network (CNN) and the Support Vector Machine (SVM). The results of the accuracy-test using the Confusion matrix method reached an accuracy value of 99.5% in identifying building similarities and 85% in classifying building types.