Research And Application Of Technology Extension Based Case-based Reasoning

China has done much work in small-town construction. Great amount of various experimental small-towns have been established in different areas, but they did not accelerate technology extension. One main reason is the successful technology application projects in experimental small-towns do not play the role of director and fugleman. If historical technology applications that have similarity and correlation with the new application problem could be found, the pace of technology extension will be speed up. Aiming at the technology extension in small-towns, it is necessary to set up a mechanism or standard of decision support in technology selection. This thesis proposes to use Case-based Reasoning theory to help decision maker choose the proper technology. Through measuring the similarity between new problem and historical cases, decision maker can acquire the part or whole solution of actual problem.This thesis introduces the difficulty of small-town construction and basic theory of technology extension, then it represents the studying status and main production of Case-based Reasoning theory and the algorithm of similarity measure. Case expression is the basis of Case-based Reasoning. This thesis analyzes the character of technology application cases referencing the sustainable development theory of small-town construction and the technology assessment, establishes the hierarchy expression model of technology case based on framework expression method and deals with the incomplete information if it occurs. Then, this thesis suggests using rough set theory and Delphi method to calculate the weight of attributes. Similarity measure is the key process of Case-based Reasoning. According to the four different types attributes in cases, this thesis uses Hamming Distance and trapezoidal function of fuzzy attributes to realize local similarity measure, and uses the Nearest Neighbor to calculate the overall similarity. The sequence according the similarities of every case framework and the whole case calculated through Nearest Neighbor algorithm represent the reference value of these historical cases. Finally, basing on similarity measure in Case-based Reasoning, a similar cases query system is designed. And it illustrates its feasibility in technology extension through a detailed example-sewage farm construction problem, and provides much help and flexibility in technology selection and extension.