Research On Multi-strategies Ontology Mapping Based On Reliability Forecasting

The goal of semantic Web is to make web data machine readable. As the number of ontologies keeps growing, ontology sharing and reuse become urgent problems. Mapping between different ontologies is the key to solve the problems of ontology communication and interoperation. The thesis focuses on research on these issues.Firstly, background of Semantic Web research domain, the purpose of researching and developing SNAX and current research situation of Ontology Mapping have been deeply introduced. Then the architecture of SNAX system and the module structure of ontology mapping sub-system LpMapping are described.Secondly, aiming to the problem existing in the multi-strategies ontology mapping system, this paper proposes a novel method of multi-strategies ontology mapping, which greatly considers the characteristics and semantic contexts of strategies. The method computes semantic similarities by making use of the ontology information of name level, structural level, attribute level, instance level and so on. So it is able to highlight those more semantically valuable information and filter those noisy or repetitive information. Then a new multi-strategy combination algorithm based on reliability forecasting is proposed.Finally, based on the research above, ontology mapping sub-system LpMapping has been designed and implemented. "benchmarks", a standard test dataset for ontology mapping provided by OAEI, is used to test the system. Experiment result shows that this system improves the precision while maintaining recall, the system thus gets a better mapping result.