2006 Volume 71 Issue 609 Pages 231-237
The purpose of this paper is to provide the method for the classification of liking for space by using Self-Organizing Maps (SOM). The preference data consisted of the replies given to a sensibility evaluation test carried out on architectural designers and students regarding several exhibition spaces. SOM includes a two-dimensional visible system which allowed us to analyze the multivariate data. By using SOM, we were able to create a map reflecting the characteristics of the exhibition spaces. In this case the pattern of the preference data was easily understood by visual analysis. Liking for space deduced from this data was classified into six categories according to the spatial characteristics of seven types of favorite exhibitions.