摘要
通过对本体进化的需求分析,提出了一个进化需求的自动生成模型。在此模型中主要通过对领域文本进行分词获得候选概念,先进行概念还原,再通过筛选、简约、转换获得最终的关键概念。采用ATF*PDF算法实现了关键概念的筛选,并引入叙词表把概念转换成规范化格式,把复合的进化需求分解成相应的原子变化,并结合相应的进化策略实现了一个本体自动进化系统。对实验的评测和分析说明了该模型能获得得良好的实验效果。
By analysing the requirement of ontology evolution,it proposed a model of ontology evolution requirement which can be generated automatically.By segmenting the domain texts,we got the candidate concepts,and got the final key concepts by restoration,extraction,conciseness and shift.This paper adopt ATF*PDF algorithm to get the key candidate concepts,and bring in the thesaurus,which is used to convert the natural concept to normalized format.Then,we decomposed the compound demands of evolution to atomic changes,and implemented an evolution system by using corresponding evolution strategy.The experiment result indicates the model can get a good result by evolution and analy-sis of the experiment.
出处
《计算机科学》
CSCD
北大核心
2013年第1期241-243,272,共4页
Computer Science