Research On Semantic-oriented Data Organization And Management Model In Dataspace

With the development of network technology and the mature of Web technology, Internet collectes a lot of information resources which are heterogeneous, evolution quickly, distribution scattered and structure loosed and so on. Traditional database technology no longer meets these new demands of data management. All of these challenges, which have brought good opportunity for the generation and development of new data organization management technique, prompt us to seek a new technology of data management. Dataspace technology is a kind of data management technology, focusing on how to satisfy new requirements of the data management.Currently, dataspace technology research concentrates on the following respects: data model, data storage and index, query processing, pay-as-you-go data integration and so on. In general it is still in the initial stage, and still there are many deficiencies and challenging work. First of all, existing data model, as the core and key technique of dataspace, has strong schema but weak content description. Second, the description of relationship among data resources in dataspace is insufficient. In order to organize and manage dataspace effectively, dataspace, as a new data management environment, to whose bottom data management must be paid attention due to complicated bottom data environment.In order to deal with the need of increasingly complex data organization and management, on the basis of the research about the current development and technology, this paper proposes semantic-oriented data organization and management model for dataspace, in which, more consideration is concentrated on the bottom of data organization as well as discovering data relationship. A perfect data description model is presented first to uniformly describe and wrap the data resources in dataspace, and then semantic relationship for the content of the metadata of data resources is constructed, which is used to recommend dataspace object and semantic entity by using Dbpedia with the rich semantic information. This model can satisfy the need of data organization and management in dataspace perfectly. The experimental results demonstrate that the proposed model and the method of semantic process are feasibility.