Research On Semantic Ontology Construction With Uncertain Data And The Application In Cloud Environment

In1982, Z.Pawlak proposed a rough sets theory to deal with uncertain data. In recent years, rough set theory and its application are developing quickly. They mainly focus on the generalization of rough set model, the uncertainty analysis in rough sets, rough set related operations and its correlation, the relation between rough set and other mathematical theories etc. Now, rough set theory has been applied in many areas, such as machine learning, decision analysis, process control, pattern recognition, data mining and so on. In the IR area, rough set is often used for expressing the uncertainty of information to expand IR into semantic retireval area. In this paper, the use of rough set in semantic retireval and semantic ontology construction will be researched.So far, there are many definitions about ontology, the most famous one is proposed by Gruber," An ontology is an explicit specification of a conceptualization". Conceptually speaking, ontologies applicated in computer domain are entities. That is a set of concepts and the relationship between them abstracted from a certain domain in the real world. The ontology technology concerns the sharing of concepts, that is the common stipulation for basic concept category in a certain domain or specific problems between intelligent agents in their interactive communicating. It is very suitable to describe the various, scattered, semi-structured information resources in Internet. By defining a shared, generic domain theory, ontologies help people communicate with computers clearly, and realize knowledge sharing and reuse between man and computer, make knowledge interaction and collaboration more convenient. Semantic ontology is the classification and structuring of the existing semantic network based on ontology theory, and with the support of ontology, it offerd semantic interoperability between information systems, as well as the intelligent access and retrieval to the internet resources.With the wide application of the semantic ontology, How to constructing semantic ontology rapidly and accurately becomes very important. As a important step in building semantic ontology, feature selection of the formal concept is essential. In the area of data mining, the importance of uncertain data is increasing. Rough set theory has been applied to text feature selection and improved many times. Also, the relation between fuzzy data and natural language has made it indispensible in semantic ontology building process. Meanwhile, with the popularity of cloud computing, it is important to consider the data placement and ontology construction in cloud environment. This paper mainly improves semantic ontology construction based on the rough set theory and other uncertain data. On the basis of that, it proposes the semantic ontology construction framework and the retrieval flow in it.The main research work includes:Combine dynamic rough set theory and Euclidean distance to improve the existing text feature selection method. Which fully consider the characteristics of text collection and the needs of users, get rid of the dependent for the decision attribute subsets in ontology construction process, and achieve a more accurate and fast feature selection.Based on the fuzzy data theory, combine hierarchical clustering and semantic ontology construction method, to improve the original rough semantic ontology construction method, proposes a new rough semantic ontology construction method, which can both consider the classification roughness and the attribute fuzziness of data.For the data placement in cloud environment, Put forward a new method of semantic ontology construction. Make a programming of the data placement in the cloud environment and make it more suitable for ontology construction.