分众分类系统中的标签通过一系列聚类算法可以形成“标签树”,但标签树中的标签间语义关系未能显性化,不能称之为标签本体。另一方面受控词表类目体系或主题词更新缓慢,跟不上网络资源新名词、新主题增长的速度,导致许多资源无法用传统分类法标引。借鉴受控词表现有的语义关系来挖掘标签树的语义关系,形成一个轻型标签本体;另一方面通过标签本体与受控词表的共享词汇,制定筛选规则,将标签本体中符合受控词表选词规则的标签纳入受控词表,使分众分类系统成为受控词表更新源泉之一,使其重新焕发活力。
Folksonomy system labels can be formed "tag tree" through a series of clustering algorithm,but the semantic relationship between tag in the tree's be missed,"tag tree"can't be called tag ontologies. On the other hand,the category system and subject of thesaurus updated slowly, failed to keep pace with the growth of new network resources,and the new theme. This had led to many resources can't use thesaurus indexing. The paper mined semantic relationships of tag tree based on thesaurus, thus built a lightweight tag ontologies; On the other hand, through a shared vocabulary of tag ontology and thesauras,making the filtering ntles,the labels those matche the selection rules of thesaurus Were chosen to into thesaurus.So folksonomy system has became a source of thesaurus vocabulary update, revitalized it.