【摘要】 针对主题图本体映射中主题上下文语义不易确定,缺少分类信息,但有明确的结构语义特征的特点,提出了A-Sim主题图本体映射算法。首先对主题图中的主题进行分类,利用描述逻辑语言ALCIR+的表达能力,根据主题属性及关联关系建立各种能表达上下文语义信息的概念术语集,将主题图本体转换为断言集合;其次在断言集合上进行多项式复杂度实例检测,记录模型构造过程中的个体,获得主题的上下文语义;最后将上下文语义结合元素级概念相似度算法计算实体综合相似度,获得异构主题图实体间的实体映射关系。通过对主题图进行相似度计算实验,结果证明新算法对主题图实体相似度计算综合性能至少提升了14%。更多还原

【Abstract】 Aiming at the problems that the context semantics of concepts are hard to make sure in the ontology mapping of topic map,and that the topics are lack of classification information but have well-defined constructive semantic characters,an A-Sim algorithm for ontology mapping of topic maps is presented in this paper.At first,the topics in topic maps are classified using expression ability of description logic language ALCIR+;then the concept terminology box expressing various context semantics of concepts is built according to the attributes and associations of topics,and the topic map ontologies are transformed into an assertions box.Second,apolynomial complexity algorithm for instance detection is presented and implemented on the assertions box,to obtain the context semantics of concepts by storing individuals in the process of constructing models.At last,the semantic similarity which combines the syntax-based and semantic similarity measurements is measured to calculate the synthetic similarity of entities,and then the mapping relationship between entities of heterogeneous topic maps is obtained.The experiments of similarity measurement of topic maps demonstrated that the novel method has achieved better performance and improved the comprehensive performance value at least 14%than other methods.更多还原