Research On Knowledge Discovery Based On Linked Data

Current Web environment evolving towards Semantic Web constantly. As the idea of Semantic Web is aLong-term vision, the Linked Data has been recommended as “the best practice” for it couldsubstantially contribute to the evolution of Semantic Web. The development and widely used of LinkedData made up the “Web of Data(Web of Linked Data)” for us. This new web resource environment hasobvious characteristics and advantages, and provided great potential for applications of AI andknowledge discovery areas. How to exploit these potential and advantages for knowledge discoveryapplication, is the crucial issues of current researches which need to be solved.This research accomplished the missions of “knowledge discovery based on Linked data”issue identifyand analysis, basic theories integration and discussion of Linked data and Knowledge discovery,resolved the process of knowledge discovery based on Linked data, then constructed the model of it.After that, this paper analysis the main functions of knowledge discovery based on Linked data and therelationship among them, build the application system of knowledge discovery based on Linked data. Atlast, designed a application system model for knowledge discovery based on Linked data and madeExperimental and verification.The issue of knowledge discovery based on Linked data caused by the demands of the development ofnetwork and knowledge activities. That’s need the integration and expansion of Linked data andKnowledge-discovery. In the framework of knowledge discovery based on Linked Data, The Linkeddata means type of data, a web resource environment, also a kind of tool for data discovery. Theknowledge discovery means general rules and basic solution for the issues, and the aim of theknowledge discovery activities based on Linked Data web. The goal of this paper is provide solution forthe knowledge discovery based on Linked data, promote the complementary between Linked Data andKnowledge Discovery.The research based on two basic theories, and combined with the general rule of science and the specialneeds of application. In this study, through the analysis of the characteristics of Linked data from basicto application, identified it’s affection on knowledge discovery, thus inferred the characteristics ofknowledge discovery based on Linked data and its potential. According these analysis, determined theprocess of knowledge discovery based on Linked data, and framed the model of it.Linked data and Knowledge discovery are both practices method, So the application system researchhas great practical significance. This study analysis the main tasks of knowledge discovery based onLinked data, then constructed the application system model. Through this system defined the levels,architecture, goal, tasks and the relationship between them.After basic rule analysis and application system research, this study designed the prototype of theknowledge discovery based on Linked data application system, then implementation and verificationaccording the prototype. The prototype design of application system based on theory analysis andapplication system research, proposed a “Schema-integration” controlled system model, better supportthe knowledge discovery activities based on Linked data. This study constructed the model of knowledge discovery based on Linked data, this model has Linkeddata as the underlying support and logic control, and knowledge discovery as the process and structurecontrol, and Link data application functions as the key operation control. This model abstracted thebasic rule of knowledge discovery based on Linked data. Also, this study analysis and framed theapplication system of knowledge discovery based on Linked data, this application system could be aunified frame for integration of existing achievements and future works. Based on these work, designedthe prototype of application system as the frame of knowledge discovery system development, andprovide reference for related works. According the results of research, this study found the testenvironment and implemented the experimental system, then test Linked data mining and associatedrule discovery with the system, verified the theory and system design of this study.This study not only extends the system of knowledge discovery theories, but also enhanced theapplication of Linked Data. The achievement of research exposed the objective rules of Linked Databased knowledge discovery, and established foundation for future work. This work promotedcooperation and fusion between Linked Data and Knowledge Discovery area, therefore promoted thedevelopment and improvement of LOD towards the Semantic-Web vision.