Creative Design Support Engine Based On Interactive Geneticalgorithm

Social development depends on the advancement of science and technology, especially the application and invention of new technologies. However, effective and innovative invention asks us to absorb the essence of previous inventions, obey the basics of invention, and exploit the unique innovative thinking of human minds. Human minds need to master a mass storage volume of specialized knowledge and yet they are restricted by fixed thinking patterns, which result in the reminded creations only in familiar fields. This is one of the biggest obstacles for innovative and successful inventions. As the development of computer science and technology, computers play an increasingly important role in innovative invention and the function of computer aided thinking has become the focus of heated discussion.This thesis firstly analyzes the current studies of innovative computing, and concludes the basic innovative computing theoretical model and its psychological basis. In the light of the fixed thinking patterns in the invention process, it designs a prototype of B/S structure -- Creative Design Support Engine based on interactive genetic algorithm, presents a detailed solution and relevant technologies, discusses the implementation methods in the process of interactive genetic algorithm, and proposes a method for the construction of the fitness function which includes the three factors of patent censoring, interactive evaluation, and machinery automatic pre-evaluation. At last, the thesis states the function of computer aided thinking of the Creative Design Support Engine through an experiment, and it is hoped that it will yield some new ideas on the part of the users. Two parts of data, the IPC and the abstract, are added into the engine, so that the patent data can be better mined and the effect of computer aided thinking can be enhanced.The original subject for the engine is the patent data, so the thesis analyzes the structural characteristics and internal semantic features in detail, and also gives a corresponding solution. To solve the extra burdens on the users in the process of interactive evaluation, the method of machine learning is introduced and the functions of automatic pre-evaluation and result clustering are realized, so that the automatization of the machine and the user experience of the system are enhanced.The Creative Design Support Engine is an attempt to further explore patent text mining research and has certain contributions to the innovative computing studies.