The Research Of Rough Relational Database Attribute Value Decomposition With Application

Relational database model was introduced by E.F.Codd in a series of his papers in 1970. Relational database theory are becoming more and more mature. However, relational data model is major used to deal with well-defined and clear data, but real-world data is often vague, incomplete and uncertain. With the enhancement of demand, many questions such as the processing of uncertainty information are exposed.Rough set theory , put forward by Z.Pawlak in 1982, is a new data analysis theory for analyzing and dealing with uncertain and incomplete data. It makes use of the equivalence relations to measure the indetermination degree of knowledge and it doesn't need any knowledge beside the data which needs to be processed. The error caused by subjective appraisal can be avoided, so the combination of rough sets and relational database to establish an expanded database model, Rough Relational Database Model is necessary, It can deal with vagueness and uncertainty data effectively, the model has obvious excellent performance.In the application of Rough Relational Database, a object's attribute has multi-values. For example, the specific attributes value is did not known of the process of data collection, but only know that the attribute may have one or several values of the value sets. In addition, in text classification, a document often can be classified to a number of categories. There many inconvenientnesses in dealing with data. In order to transform the actual operation of Rough Relational Database, the thesis put forward the theory of attribute value decomposition.In this thesis, the concepts of rough set, rough relational database, attribute value decomposition algorithm and the elementary systemic application of the model in OLAP are discussed. The mainly work are in this paper as following:(1) The concept of single-valued attribute, Single-valued attribute decomposition, decomposition theorem in rough relational database model are proposed in this paper.(2) Given the order for multi-value attribute's values of rough relation, wiht rough relation entropy, the value of multi-value attribute are choosed based demand. By decomposing the Rough relation, standard relation is obtained.(3) A attribute value decomposition algorithm is proposed, the time complexity and the properties of the algorithm are discussed in detail, the algorithm's effectiveness and feasibility are verified by an example.(4) Given the research of the single-value decomposition, the thinking of multi-value decomposition propose is involved.(5) Improving the traditional architecture of OLAP, the result of the attribute decomposition in OLAP is applied. It is enhancing positioning accuracy of the information, and improving the efficiency of data analysis tools with respect to analyze the rough relation directly.At the end of this paper, contents are summarized and some further works are proposed.The main innovation selects in the paper are as follows:(1) The feasibility of attribute decomposition is proved, and the concept of single-valued attribute, single-valued attribute decomposition, decomposition theorem in rough relational database model are proposed .(2) The importance of rough relation's attribute value is determined by rough relation entropy. It is improving the accuracy of information uncertainty with respect to roughness. The decomposition algorithm's feasibility is verified by an example's code.