設計思考における要素の導入について : 建築知識工学的観点から

Abstract

I define design process is not goods-product process but knowledge-product process. Namely architectural value is added by the accumulation of knowledges, rather than by the accumulation of mere man-power/day. What is a knowledge? First of all, the knowledge must be written in a sentence or sentences. The things in mind is not a knowledge. Those sentences must be understood by a computer system running on a certain computer language, here on Prolog. Understanding by computer means that the computer system can compute and answer by 'yes' or 'no' (or corresponding form to it), if the sentence is input by a certain form, here by logical formula. If the answer is 'no', we have to compliment the first input sentence by adding suitable sentences. Then if the computer succeeds to compute the sentence by answering 'yes', the initial input sentence become a "knowledge". If the computer system stores all of inputs in its data base for the convenience of later referring by someone, namely a user or other computer system, I say the knowledge is externalized. I discussed and descriminated three levels of knowledge, namely 'information level', 'semi-knowledge level', and 'knowledge level' (Chap. 2). 1) Information level '. This means a sentence is not articulated yet and remains as a vague whole. 2) Semi-knowledge level.' This means a sentence (or its logical formula) is articulated but has not been satisfied by a set of elements yet. 3) Knowledge level; This means a sentence (or its logical formula) is to have been satisfied by a set of elements. So it is said in designing the knowledge comes first as a semi-knowledge and becomes a knowledge or remains a semi-knowledge. This is materialised as the condition that the translated logical formula of a natural language sentence includes only variables and no individual constants. In computing the logical formula, an element or pairs of elements are given by a user (designer) or from data base to a variable as instances (or values) which satisfy it. In other words, computing is to make a semi-knowledge a knowledge by giving or generating instances. But note that computing on Prolog, it is allowable for variables to remain variables if they are included in facts on the data base. So the answer 'yes' by Prolog computer system is not necessarily to show all variables are satisfied by some instances. The explication of computing as a recursive function by Kleene is that it has two phases, one is to generate an element as existent and the other is to introduce a symbol to refer to the element, and that these two phases must have the same order. It, as a matter of course, corresponds to computing by Prolog as later shown in Fig. 5 (Chap. 3, 4). Design thinking has a certain correspondence to computing if the design theme is given in a simple or complex logical formula and is sastisfied by being given elements to all parts of logical formula which the designer in-tends to materialize. Therefore the design process can be followed by computing process through the interaction between the designer and the computer. There comes the design results as the accumulation of input data by the designer in the computer's data base. The computing process of a logical formula (5-4) which is shown diagrammatically is Fig. 5, where the bare number is the order of computing and the number with * shows the point where an element is logically introduced in computing. The arrow shows the success of correspondence in a sense, namely matching in Prolog. Fig. 5 is the computing process pattern of a particular sentence and if input sentence is changed, the programme may be changed, but 1 have developed more complex program which can compute most of logical formula, Let us correspond Fig. 5 to the design process. If we draw Fig. 6, this is completely circular and nothing