Research And Application On Design Knowledge Personalized Search Technology Based On Semantic

With the development of information technology and the accelerated process of economic globalization, the upgrading of new products is becoming faster and faster. The efficient product design is the key to enhance the market competitiveness of enterprises. Knowledge management system enables designers to store, retrieve, integrate and update the design knowledge effectively during the process, to achieve the goal of knowledge, continuous accumulation and innovation. In the meanwhile, the designer’s knowledge becomes the intellectual capital of the enterprises, which may contribute to the design inheritance and innovation, and speed up the market-demand response. In this paper, personalized search based on semantic will be introduced in the product design knowledge system, based on the business needs for design knowledge management. The main contents of this paper include:(1) Building a framework based on semantic and personalized search, namely, the IR system and personalized search agent are introduced on the application layers to complete the semantic expression of design knowledge, fulfill the users’personalized needs, and also enable the system to have good encapsulation and service.(2) Proposing a construction method for semantic design knowledge document model. This paper improves the traditional calculation of feature weighting to enhance the expressing ability of the feature words; building the design knowledge network by borrowing its concept simultaneously, and use it as a semantic understanding components; Finally, proposing a similarity calculation method for design document model.(3) A user model is given based on intelligent adjustment algorithm. To solve the personalized information agent service problems, the paper combined custom modeling with automatic analysis of hand-modeling technology, mining user behavior log on depth and to improve the expression of user interests, to track changes in user interest features, combined with page feedback calculation from the user browsing to complete the adaptive user interest model.(4) We completed the design and development of truck design knowledge management system. We achieved knowledge express, storage and personalized retrieval during the design process. The practical use shows that knowledge retrieval precision and recall are improved obviously, our method greatly enhances the design quality of their products.