Font Size: a A A

Research Of Migrating Relational Database Based Legacy System To Semantic Web

Posted on:2011-01-08Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y X ChenFull Text:PDF
GTID:1118330332478372Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
After years of development, the IT department of a large company produced a lot of systems, which were based on different architectures and without any unified planning. If the distributed knowledge in these systems can be integrated and managed effectively, the enterprise can setup value-added services and benefit a lot from them. Since most of the information systems running on top of relational databases, one idea is integrating the relational databases to the semantic web layer of the corporation, and migrating the legacy information systems onto the new data layer.In this dissertation, the relational database based information systems and the techniques of Semantic Web are analyzed. Some problems in the migration from relational database based legacy system to Semantic Web are discussed. The main research work and contributions of this dissertation are as follows.(1) One ontology learning approach is selected and analyzed, and then verifies that SPARQL is relational completeness in the scenario.Therefor, the SQL queries in the legacy system could be replaced by SPARQL queries in the migrated system. Thus, those old but useful systems which are based on relational databases can be migrated to Semantic Web Applications.(2) An automatic approach of queries transforming from SQL to SPARQL is researched. Based on the mapping functions between Ontology and database, the SPARQL algebraic and Construct expressions are used to simulate five basic relational algebra operations.The given SPARQL algebra expressions can be composed to simulate more complex relational algebra expressions. So the basic SQL statements can be transformed to SPARQL queries automatically.(3)A Semantic-Aware concept ranking algorithm is proposed for OWL Ontology. This algorithm abstracts semantic meanings of the edges from Ontology graph, and considers the semantic correct paths for the iterative flooding algorithm. The flooding algorithm can combine with the user's feedback to increase the ranking precision, and help understand the complex Ontology learned from Database.(4) The injection attacks in Semantic Web are studied, and a prevention approach is proposed.It presents the possible injection attacks in RDBMS based system on the Semantic Web, and gives a classification of them. An injection detecting approach is described to protect the migrated Semantic Web Applications from injection attacks.Finally, a financial legacy system is migrated under above technology. The new system could provide the same functionalities, but is more flexible for future enhancement. Our work enriches the research of ontology approaches applied in software migration area. The study results of the thesis benefit to further research in migrating software to Semantic Web.
Keywords/Search Tags:Semantic Web, Ontology, system migration, Ontology learning, relation query, SQL, SPARQL, concepts ranking, injection attack
PDF Full Text Request
Related items