| Cloud Platform of Agricultural Technology Extension(CPATE) is a comprehensive knowledge platform serving for the agricultural technicians all over China.It provides various knowledge services needed by agricultural technology extension.Along with the development of the project, the data quantity increases rapidly.The platform urgently needs an effective tool to help the technicians in finding the most needed knowledge at short notice and to ensure the knowledge being learned and internalized.Knowledge map, as a knowledge management tool, has been applied in many fields.These maps will provide users an intuitive and graphical knowledge guide, reveal the relationship between knowledge, and promote knowledge exchange and sharing.Based on the demand of CPATE and the characteristics of knowledge map, this paper puts up an idea about applying knowledge map to the field of agricultural science, and designs a scheme of constructing knowledge map in CPATE. The scheme has been implemented partially and the effectiveness of the core algorithm is verified. This paper focuses on the following researches:1. Analysis of the current situation of knowledge map research and identification of its development trend are accomplished, and characteristics of knowledge map is summarized. The type of knowledge map of CPATE is differentiated. Classification and summary of the construction principle and method proposed in the past research is accomplished.2. This paper introduces the basic situation of CPATE, and analyzes the demand and effect of the platform. The characteristics of users in the platform is deeply studied, which find that they have sufficient information processing ability, but the ability of information retrieval is weak. The knowledge resources of the cloud platform are analyzed in detail from three aspects as knowledge generation, knowledge content and the structure of knowledge.3. The research of key technologies for the construction of knowledge map on CPATE is illustrated in this paper. The technology of knowledge nodes generation is studied, including the Chinese word segmentation technology and social tagging technology; a suitable nodes correlation algorithm is selected and improved according to the demand of the platform by comparing popular correlation algorithms; the vector space model and the latent semantic indexing model are studied in detail; the visualization technology of knowledge map is studied. Since the cloud platform focus on mobility, the mobile display technology is analyzed.4. The construction plan of knowledge map on CPATE is designed. According to characteristics of the users and the platform, the construction principle of the knowledge map is determined; according to the demands of cloud platform, the function of knowledge map is designed as knowledge guidance, knowledge recommendation and providing interface for data analysis; the total structure of knowledge map is designed as six levels, and the detail design is explained, including the production of knowledge nodes, correlation algorithm of knowledge network and the display scheme of the knowledge map.5. The construction scheme of knowledge map is implemented partially; the effectiveness of core algorithm is verified and the result is acceptable. At last, the strategy and design of displaying knowledge map are illustrated in detail. |