As with the last paper, here we focus on the cognitive map that expresses the interaction between architectural/urban space and human beings in order to develop a design method for the Human-Environment System in which all the elements including human beings continue to interact with each other. Sketch maps drawn by university students are analyzed as externalized drawings of the cognitive map as with the previous paper.
This paper proposes a method to evaluate co-occurrence probabilities between architectural signs on a sketch map via logistic regression analysis and analyze the disposition of the architectural signs with their probabilities on a geographical map. A simulator created as below is used to identify the impact of the change of some characteristics of the architectural signs on their co-occurrence probabilities, considering their application in Human-Environment System design as follows.
First, by focusing on the meaning aspects of the abovementioned interactions, the cognitive mapping process is understood as a thinking process through signs on the basis of the concept of semiosis proposed by C.S. Peirce. Therefore, the cognitive map is modeled from the interpretation of the architectural signs that represent the architectural/urban space using the concept of “architectural sign, ” “architectural/urban space, ” and “cognitive map, ” following Peirce's concept of “sign, ” “object, ” and “interpretant.” On this basis, the “co-occurrence” between two co-occurring architectural signs is understood as the relationship between “sign” and “object.” Then co-occurrence types of all the pairs of architectural signs are classified into three categories based on his concept of “icon, ” “index, ” and “symbol.”
We then propose variables for the relationships among the co-occurring architectural signs to determine these signs' co-occurrence types. Then using the GIS database created from the drawn sketch map, the logistic functions are obtained via logistic regression analysis, which uses the proposed variables as independent variables. From these logistic functions, we can predict in which co-occurrence type the co-occurrence probabilities will be higher (or lower).
Then we create a simulator on which architectural signs' dispositions with the co-occurrence probabilities are drawn on the geographical map using the obtained logistic functions. Thus, we analyze the impact of the change of some characteristics of architectural signs on the disposition of their co-occurrence probabilities on the geographical map.
Each university is found to have a different co-occurrence tendency depending on the co-occurrence types of its architectural signs through the analysis. For example, in Matsuoka campus, which is located far from the center of the city and whose students use their own cars every day, the co-occurrence tendency of an architectural sign located along the bypass far from the university is unaffected by the change to architectural signs near the university; however, it is affected by changes to architectural signs along the bypass. In addition, some architectural signs co-occur with many others, such as those for convenience stores or supermarkets, the change of whose architectural characteristics can have a major impact on the co-occurrence tendency of the targeted area.
The analysis method and results presented here show the possibility of considering the impact of architectural design on the cognitive map of the targeted area, which indicates a way forward for the Human-Environment System design.
抄録全体を表示