Research On Knowledge Fusion Algorithm Of Multi - Source And Heterogeneous National Information Resources

As an interdiscipline, knowledge fusion has been used in various fields. Atpresent, most Internet service just limited to data integration on the data layer, whichcannot well reflect the interaction between the information as well as the meaning ofcollaborative. While it can achieve high-quality services by generating newknowledge on the basis of traditional knowledge in the process of designingknowledge integration rationally. There are abundant resources of ethnic minoritiesin Yunnan province. These resources have their own trait of heterogeneity anddisorder, which leads to that fewer knows each ethnic characteristics of theinformation. This paper aims at the multi-source and heterogeneous character of thenational information resources and analyzes from the perspective of knowledgeintegration. Relying on the correlational research of Ministry of Education KeyLaboratory of National Education Information Technology, it gives a knowledgeintegration model of the multi-source heterogeneous national information, and eachmodule in the model has been analyzed and implemented.This paper is mainly to solve the following aspects:(1)Knowledge Fusion is a re-organization process based on the knowledge base.The fundamental task of knowledge fusion of national information is to provide theuser with the transparency, consistency and real-time performance of a variety ofheterogeneous ethnic information. The paper aiming at this point provides ageneration model on multi-source heterogeneous national information resource anddiscusses the build process of knowledge.(2)The target of knowledge fusion model is to generate an effective andcompleted knowledge repository. The paper combining knowledge metatheory,discusses it in detail on three parts, pretreatment process, similarity calculation aswell as fusion rules and algorithms of knowledge element. This part is the theoreticalfoundation of knowledge fusion model.(3)The key point of knowledge fusion model is to design a reasonable knowledge fusion algorithm. The paper constructs knowledge fusion algorithm KFAbased on genetic algorithm theory. To further optimize the solution space, relying onAnt Colony Optimization algorithm, the paper provides a knowledge fusionalgorithm KFA-ACO based on user demand. Meanwhile, the paper has described theabove algorithm in detail and analyzes the algorithms combining with simulationexperiments.(4)Paper points out the key issue of knowledge fusion in knowledgevisualization systems needs to be solved seriously. At the same time, it also gives avisualization prototype system model of multi-source heterogeneous informationresources based on knowledge fusion and has been described and implemented eachmodule in prototype system.