Research On Bigraph Model-based Large-scale Ontology Partition And Mapping

Ontology mapping aims to build a semantic bridge between heterogeneous ontologies, through which the goal of knowledge sharing and information exchanging between different ontologies can be achieved. However,with the continous development of the ontology application,the size of ontology is also growing. Then,the large-scale ontology mapping has become the hot spots in the current semantic web research.This thesis will focus on the large-scale ontology mapping research.Firstly, the research background and significance of the thesis is briefly introduced, then, the current large-scale ontology partition and mapping methods are summarized. After, we have analyzed the problems of these methods,and pointed out the direction for our work.Secondly, drawing on the idea of graph theory, we propose a new method for large-scale ontology partition and mapping based on bigraph model. Compared to traditional partition method, our method uses the idea of co-partition to partition the two ontologies synchronously and achive block matching in partition phase. It first computes the conceptual similarity which contains text similarity and lexical similarity between the two ontologies. Then, a weighted bigraph is built by the conceptual similarity. Now, the bi-partition problem is translated into a graph partition problem. After, we find the1:1mapping in the subgraphs by optimal matching algorithm, and apply the revised rules to improve the mapping quality in the optimal matching.Thirdly, against the mapping which obtained by the similarity lack of semantic background knowledge, we apply the revised rules to revise the candidate mapping set. It through the discovery of the missing mapping and eliminate conflict mapping to obtain the final semantic mapping results.Finally, based on the research above, Bigraph based Partition and Mapping (BGPM)system has been designed and implemented. Moerover, the mapping effective of BGPM has been tested through using ontology set offered by OAEI compared with the other large-scale ontology mapping system. The experimental results show that our method can acquire high partition quality and good mapping efficiency. It also can find a variety of mapping types.