Research On Semantic Heterogeneity Recognition And Elimination Of Multi-Source Network Information

Heterogeneous, caused by multi-source, is the essential feature of network information:the difference of ideas and methods that expressed by cognitive subjects who hold different worldviews during the process of information collection and processing leads to various levels and types of heterogeneity such as system heterogeneity, structure heterogeneity, syntax heterogeneity and semantic heterogeneity. With the development of research for multi-source heterogeneity, the focus of researchers for heterogeneity is shifting from technical to cognitive level. Thus, semantic heterogeneity has gradually become the focus and difficulty of automatic processing of network information.Price of agricultural product, important agricultural network information, is one of the most typical multi-source heterogeneous information. In fact, multi-source heterogeneous in agricultural product price that collected by the Ministry of Agriculture and various wholesale markets can be eliminated by manual processing. So how to achieve the automatic processing of information by using the technology of eliminating multi-source network information is a key issue of efficient utilization of agricultural informationSeveral indicators, such as network price information distribution mode, information type, the frequency of updates, price information attribute value and so on, are collected by using the method of network research. With these indicators, we extract some common characteristics of the multi-source network information of agricultural product price and describe them with nine key properties. Drawing on the design ideas existing in the theory of information life circle to backtrack from heterogeneous appearance to the information collection programs to explore the generating reasons and external characteristics of different types of heterogeneity, this dissertation has formed the characteristics sets of different semantic heterogeneity. According to the hierarchical theory of information processing, we analyzed and put forward three levels of semantic heterogeneity, which are semantic heterogeneity based on schematic, context data and abnormal individual data, to propose eleven kinds of more detailed classification framework of semantic heterogeneity about agricultural product price information in multi-source network. Based on the hierarchical relationships of semantic heterogeneity, a method to eliminating semantic heterogeneity whose main content is ontology mapping is presented by using the semantic heterogeneity based on schematic as an example. At the same time, we put forward the overall framework to eliminate semantic heterogeneity of agricultural product price information in multi-source network, and then design a prototype system framework.Through studying the formation of semantic heterogeneity of agricultural product price information in multi-source network in a cognitive perspective, we proposes the corresponding framework of semantic heterogeneous classification and elimination by using the theory of information life circle and the hierarchical theory of information processing to analysis these types and characteristics of semantic heterogeneous phenomena. In a sense, this method has some innovation on research.