Research On Technique Of Semantic Query In RDBMS

With the rapid development of network technique, the size and complexity of huge information resource is now growing with the index grade, so it set higher requirements for intelligent management of huge information resource, and must solve the key technique about information respresentation, information storage and semantic query based on ontology, and so the semantic query in RDBMS is of special value.Based on specialized research management of huge information resource produced by open network, this thesis is intended to contribute to key technique of semantic query in RDBMS, and mainly involves the following issues.1. Propose the model of semantic query in RDBMSAccording to the characters of semantic query on huge information resource, we propose the module of semantic query in RDBMS. The model consists of three key components. The first part is about ontology technique, and it is the basis and foundation of semantic query. In this part, we express the building and storage strategy for ontology. The second part expresses information respresentation and storage. This part includes information respresentation, semantic annotation and information storage. The last part is about semantic query. In this part, we express how to implement semantic query in RDBMS, and this part is the kernel of the research.2. Building concept semantic similarity computational modelThe measurement of concept's similarity is basis and foundation of semantic query based on ontology. In this thesis, we analyze the meaning of similarity and relativity. On the basis of psychological studies about similarity and artificial intelligence studies about similarity, we propose a model to measure the semantic similarity between concepts.3. Introducing the measurement model of concept's similarity into RDBMSWith the query process of RDBMS, take measurement of concept's similarity as a reason model, and introduce it into the RDBMS (?)architecture. This model provide reason interface for ontology storage and concept's similarity measurement. 4. Adaptability adjustment of center system architectureAccording to the characters of center system architecture, we analyze the adaptability adjustment of the system in order to support semantic query, and then propose the new simplified model of center system architecture, at last, analyze the adaptability adjustment of the all module in the system.5. Propose structure join index based on node relationshipOn the basis of the characters of shredding storage of XML data, propose a new index strategy to improve the query efficiency. The information between the nodes is not direct after the data is shredding into the table, and so it has to some strategy to keep the information. In the node relationship algorithms, an XML document can be transferred an order tree, tree nodes correspond to the document elements, and edges correspond to the element-sub element relationship. In the relationship join algorithms, element index and node relationship index are the two indexes to support the node relationship joins. This relationship index has five search keys, XML document id (DID), node id (NID), his parent node id (PID), his first right sibling (SID), his first left child node id (cid). The relationship index helps us to determine the relationship of two nodes quickly to speed up the structure join query.