Research On XML Conceptual Model And Internal Transformation Method

XML is widely used in the Internet data transmission, it has been regard as a dataexchange and storage standard in many cross-platform application system. XML has manyadvantages, such as self-description and expansibility. In order to make the XML candescribe the object in the real life better, and all kinds of relationship between them, it isnecessary to use a conceptual model to the modeling. A conceptual model can abstractvarious things in the real world into the form that information system can identify, and thenplay an important role in the field of computer. A well-form conceptual model can not onlyembody the characteristics of XML itself, but also lay a solid foundation for further study ofthe application of XML.In this paper we analyzed the research of XML conceptual model, including presentsituation, the existing problems and improving methods, based on the research at home andabroad. To the question that extended E-R model failed to describe the hierarchic structureof XML document, we proposed a double level conceptual model of XML. In this model wekept the advantages of E-R model, and also introduced hierarchic structure to the conceptualmodel, so that it can reflect the hierarchical structure of XML better. In the firt level ofdouble level conceptual model, we made some modifications for some basic E-R conceptssuch as entity and so on, in order to make adaption for the characteristics of XML. We alsoput forward the concepts of data node type, outgoing clustertype, incoming cluseter type,etc. At the mean time we gave the formalized hierarchic mapping, and add order into theintegrity constrains, make it more suitable to the XML document. The main function of thesecond level is to transform some parts of the first level into the hierarchic structure.And then we put forward the transformation algorithm from the global concept level tothe hierarchical level. There are corresponding relations between the two levels. We canobtain the hierarchal structure from global concept level by the transformming operation.First of all, we selected relationship type and weak entity type from the overall conceptstructure. There are assigned transform algorithms for these selected types. Then, we usejoin operarion to combine these hierarchical structures together. In some particularsituations, there may be contraction occurs among hierarchical relationship types, at this point, their father and children may be merged into a hierarchical entity type.