Spatial Analysis of Chinese Garden Designs with Machine Learning

Abstract

This research intends to propose a scheme for analyzing Chinese Garden Design by incorporating spatial theory, data mining, concept of object, and network-like data structure. Design elements of Chinese garden are placed in a network according to the existing gardens according to spatial theory. Collected networks are then divided into pair of elements connected by their relationship and stored in a database. Later, data mining is applied to attain patterns from the node-and-relationship pairs. Meanwhile, the elements of the same level can be classified and data grouping can be done by the implementation itself. Thru this research, we can gain insight upon the spatial information and relationship between elements of Chinese garden designs. The result is a set of more concise and structural descriptions, which reveals the rhythm behind the Chinese garden design and can be a great pedagogical aid.