Method Of Extraction Of Inspiration Knowledge In Patent And Product Innovation Based On Memory Structure

After analyzing the recent development of research of patents,design fixation and inspiration,this paper proposed the method of patent screen and the means to extract inspiration knowledge in patents,and developed the product innovation design system based on patent knowledge,for improving the current retrieval method of patent text and the searching efficiency of relative patent knowledge in the process of product design.There are five chapters in this paper:The first chapter illustrates the research background and meaning of our topic,mainly introducing the recent development of relative studies and determining the research direction and main contents.In the second chapter,this paper proposed the retrieval method of patent text based on Topic Model in view of the deficiency of current method such as the way of using vector space constructed by word frequency,making use of target patent to screen the patents again after the filter through keywords.In the last of this chapter,a test was made by using American patents for verifying the effect of the method proposed in this chapter and testing the effect of using different parts of patent text.The third chapter studied the methods of extracting and pushing knowledge in patents from the perspective of human memory structure,thinking pattern and experience of different designers.At last,the chapter described the process of the invention of the device to decelerate and control the motor in circuit breaker to test the effect of the method proposed in this chapter.The main content of the fourth chapter is to introduce the Computer Aided Product Innovation Design System based on Patent Knowledge,which is developed based on the theories illustrated in the chapters mentioned above.This chapter described the functions of different modules in this system and verified the efficiency through the introduction of national invention patent produced by this system.The last chapter summarized the work in this paper and pointed out the following work in the future.