*2.1. Urban Morphological Quantification*

Multiple quantitative approaches have been implemented for transmitting morphology to data. Researchers have used pre-proposed indicators (e.g., semantic indicators and geometrical indicators) to index buildings and urban relationships for the following studies [20]. For example, in the early stage application of case-based reasoning for design, ARCHIE is an intelligent case browsing system. The cases are represented as attribute-value pairs with 150 features, including the concept, text, actual number, and function [21]. In the study of urban typologies [22], block and street types were studied as a context-sensitive sample of types by describing the cases with semantic and geometric values, such as land use, length, area, and ground space index (GSI). Nahyun developed a model adopting case-based reasoning and a genetic algorithm to predict the maintenance costs for aging residential buildings [8].

On the other hand, hierarchical structures along with symbolic representations have been constructed to describe urban forms containing plots, buildings, and streets [23,24]. Based on the established hierarchical structure, measurements (e.g., distance and connectivity) were calculated for further analysis. For example, in the study of spatial design network analysis (sDNA), Crispin developed a tool for network analysis, which was implemented for the representation and calculation of the network nodes and link density [25].

Song proposed an access structure to symbolically describe and measure the relations between the fundamental elements: plots, buildings, and streets. They took eight cases for similarity analysis to validate the access structure [26]. Although discrete indicators and structural measurements help in morphology quantification, there are still features that are hard to describe numerically—especially the geometric information, such as the volume distribution, directions, shapes, and connections. This leads to a generalizability limitation.
