An automated approach to scoring architectural designs of building sites was devised. Design drawings were represented in the computer as a database consisting of a series of objects and their locations and relationships to other objects. Designs were automatically scored by: (a) determining whether or not certain basic requirements were met (such as parking, handicapped access, and building within site boundaries) and (b) determining the extent to which the drawings met criteria of efficient site design. The drawings had been scored earlier by a panel of expert registered architects. The automated technique was able to reproduce the human jurors' ratings for many drawings. In particular, drawings that the automated technique failed were quite likely to have been failed by human jurors as well.
[1]
Ö. Akin.
Psychology of architectural design
,
1986
.
[2]
U. Flemming,et al.
A Generative Expert System for the Design of Building Layouts
,
1986
.
[3]
Bill Hillier,et al.
Ideas are in Things: An Application of the Space Syntax Method to Discovering House Genotypes
,
1987
.
[4]
Garry Stevens,et al.
Cadd Made Easy: A Comprehensive Guide for Architects and Designers
,
1987
.
[5]
Isaac I. Bejar,et al.
A methodology for scoring open-ended architectural design problems.
,
1991
.
[6]
Jacob Cohen.
A Coefficient of Agreement for Nominal Scales
,
1960
.