A QUANTIFICATION ANALYSIS OF CAMPUS BEHAVIORAL DYNAMICS USING THE INFORMATION ENTROPY

Abstract

Measuring complexity of the dynamic system has become a common practice for describing spatial structural properties in the fields of urban geography and landscape ecology. In China, college campuses can be regarded as a kind of complex system since the campuses accommodate multiple functions, such as education, research, leisure, residence and so on. Considerably diversified human activities are daily performed in campus open spaces. How to characterize the distribution dynamics of daily activities calls for much attention of architects and planners. Nonetheless, the resultant dynamics of human activities is often irregular and patchy, giving rise to intricate distribution patterns that can be difficult to characterize. Herein, the issue of characterizing the temporal-spatial-categorial Diversity of Activity Distribution (DAD) in open spaces was addressed and a method of quantifying the complexity of patchy activity dynamics was proposed. The method was inspired by information-based measures of entropy, and the proposed Behavioral Entropy Index (BEI) can distinguish the distribution of activities in open spaces between simple (convergent) and complex (random) temporal-spatial-categorial mosaics. The method was demonstrated using sample data through a survey on two typical college campuses at Hangzhou City, China. The results show that the BEIs effectively illuminate the behavioral dynamics, rather than the conventional index of absolute population or simple percentage; moreover, proper facilities, natural environments and campus management all facilitate improving the behavioral complexity.