基于参考点的大规模本体分块与映射

针对大规模本体映射中存在查全率和查准率不高的问题,提出了一种新的基于参考点的大规模本体分块与映射的方法.该方法的主要思想是用参考点来指导分块,并同时对待映射的两个大规模本体同时分块,即联合分块.首先对大规模本体进行预处理,将本体中的实体名称归一化并将其表示成本体树的形式,然后采用一些简便的方法找到参考点,最后以参考点为聚类中心对两个本体树的概念进行聚类,并同时实现块映射.理论分析和实验结果表明,该方法能够有效地解决大规模本体映射问题,并能获得较好的查全率和查准率.

Anchor-based large-scale ontologies partitioning and mapping

In order to solve the problem of low precision and low recall of large-scale ontology partitioning and mapping, this paper proposed a new anchor-based large-scale ontology partitioning and mapping method. This method used anchors to guide partitioning, and partitioned the two ontologies at the same time, which called co-clustering. Firstly, it preprocessed the two ontologies in order to normalize the entities's name and turn them into tree structure, then used some simple methods to find anchors. At last, the anchors acted cluster centers to cluster the concepts in both ontology trees, and found block mappings at the same time. Theoretical analysis and experimental results show that this method both solves the large-scale ontologeis mapping problem and achieves good precision and recall.