Font Size: a A A

Research And Application Of Data Model Based On Description Logic

Posted on:2017-12-09Degree:MasterType:Thesis
Country:ChinaCandidate:H F LiuFull Text:PDF
GTID:2348330533950179Subject:Computer technology
Abstract/Summary:PDF Full Text Request
Model Driven Architecture(MDA) is a software framework developed by the Object Management Group(OMG) to standardized software development process raised. In Model Driven Architecture, model is regarded as a product of the software development process, all processes of the software development process can be regarded as modeling. In database design, the model is data model. A good data model can reflect the structure, operation and constraints of the data as much as possible. Currently the most widely used of the data model is UML class diagrams,the graphical way what the UML class diagram used is convenient and intuitive, but it has some disadvantages. First, UML class diagram graphical model method does not have a precise semantics, thatmay cause ambiguity. Secondly, UML class diagram does not have the formal characteristics, so that the model is difficult to reason by computer. Description Logic is a formal knowledge representation tool, it has a clear model-theory mechanism, which can reason the results Description Logic represents. Based on this feature, many scholars represents the data model by Description Logic, to check the semantic error of the data model.Data warehouse is modeled by MultidimensionalModel. Sometimes the data can not be aggregated rightly because the MultidimensionalModel has design flaws. Usually, the designer check the summarablilty and consistency of the MultidimensionalModel after modeling MultidimensionalModel. Checking MultidimensionalModel by people not noly easy to make mistakes, but also slowly. In order to use computer to check consistency and summarability of the MultidimensionalModel automatically, this thesis presents a method to check model based on Description Logic. Firstly, transport the meta-model of the MultidimensionalModel to Description Logic SHIQ knowledge base, then analyze the semantic between meta-model and model, get the conclution that using description logic SHIQ can represent MultidimensionalModel rightly. Secondly, analysis the consistency and summarability of MultidimensionalModel, provide a method to check consistency and summarability of MultidimensionalModel by Description Logic SHIQ. Finally, use this method in telecommunications MultidimensionalModel, and use RacerPro to check the consistency and summarability of telecommunications MultidimensionalModel.
Keywords/Search Tags:Knowledge representation, Description logic, Data warehouse, Data model
PDF Full Text Request
Related items