A boundary knowledge field based data mining method

Shan Feng

School ofComputer Science and Technology, Hubei Polytechnic University, China

Data mining (the analysis step of the "Knowledge Discovery in Databases" process, or KDD), an interdisciplinary subfield of computer science, is the computational process of discovering patterns in large data sets involving methods at the intersection of artificial intelligence, machine learning, statistics, and database systems. Knowledge is the source of getting and keeping the strength of competence, knowledge becomes an important strategic resource at the age of knowledge economy. The way to facilitate knowledge transfer smoothly plays a critical role in determining the competence of an organization or organizational system. To explore the mechanism of knowledge transfer, the characteristics of sticky transfer (flow) is compared with the flow characteristics of viscous fluid, and the knowledge on the theories of field in physics, viscous fluid mechanics and boundary layer is employed and analysed in this paper. Firstly, the concept of boundary layer in the knowledge field is proposed to analyse the difference of knowledge stickiness in the knowledge field and describe the contradictory relationship between knowledge stickiness and liquidity. Secondly, based on the analysis of the knowledge transfer in the boundary layer, the dynamic mechanisms of sticky knowledge transfer are analysed from the three aspects of the knowledge potential difference force, the viscous force and the extern driving force. The rotation mechanism of knowledge in the knowledge field is discussed, and the dynamic model of the boundary layer of knowledge field is built. Finally, the phenomenon of knowledge flowing into and flowing out of boundary layer is discussed, and the knowledge transfer conservation equation in the boundary layer is constructed to describe the updating and appreciation of knowledge in the boundary layer of knowledge field.