基于最小信息熵分类的不确定元数据本体构建

MIE-categorized uncertain metadata ontology construction

RDF中的元数据具有无序性的特点,使得获取所需元数据时查全率较低,且由于获取的RDF可能存在不确定性问题,造成查准率较低。针对以上两个问题提出:利用概率公式,将不确定RDF抽象为初始化模糊概念集,利用最小信息熵理论做模糊等价分类,利用粗糙集理论,计算得到粗糙模糊概念;利用概念格将粗糙模糊概念建立偏序关系,得到一个有序的层次关系模型并用OWL语言表示。经实验验证,该方法正确有效,可以在原方法基础上提高查准率和查全率。

In RDF,metadata are disordered,which makes the query have lower recall ratio when acquiring the required RDF to get its data.Meanwhile,because the uncertainty may exist in obtaining RDF,precision ratio is lower in use.To solve above two problems,two works were done.The uncertain RDF triples were abstracted as an initial fuzzy concept set using probability formula,minimum information entropy(MIE)theory was employed to do fuzzy equivalence classification,and the classified concepts were treated as rough formal concepts through rough set theory.The partial order relations from fuzzy rough concepts were constructed using concept lattice,and an ordered-hierarchical relation model depicted in OWL was set up.Experimental results show that the proposed method is correct and effective,and the precision and recall rates were improved compared to the original method.