摘要
提出了一种估计本体包容性(comprehensiveness)的方法。与一般方法不同,该方法不是简单地把本体中类和属性的个数当成本体的包容性,而是通过对语料库的统计分析,发掘出本体词汇在自然语言中的出现频率,从而更准确地反映了本体的包容性。它考虑到了本体的内容和包容性的关系,是对本体更深入地分析,因而对本体的构建和选择具有更好的指导意义。
A novel method is proposed to evaluate the comprehensiveness of ontology. Comparing with other methods that take the number of classes and properties in ontology as comprehensiveness measure, this method reveals the commonness of vocabulary in natural language by applying statistical analysis to a corpus, so as to capture the comprehensiveness of ontology more accurately. This method concerns about the connections between the content representation and comprehensiveness of ontology, which is expected to have a better guidance for ontology constrnction and selection because it renders a deeper analysis into ontology.
出处
《计算机科学与探索》
CSCD
2009年第6期633-640,共8页
Journal of Frontiers of Computer Science and Technology
基金
国家自然科学基金
国家重点基础研究发展规划(973)
江苏省自然科学基金项目~~
关键词
本体
评价
包容性
度量
ontology
evaluate
comprehensiveness
metric
