欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
通过分析已有的中文本体概念层次获取方法的特点和不足,提出一种利用度属性获取概念层次的方法.首先将概念关系以图的形式表示出来,然后利用反向剪枝算法得出概念所在层次,最后补充并修剪层次关系生成完整的概念层次.从各种参数的对比结果来看,基于VSM（Vector Space Model）且不进行LSI（Latent Semantic Indexing）降维时效果最好,同时也优于目前已有的几种方法.
英文摘要：
By analyzing the characteristics and disadvantages of several Chinese ontology concept hierarchy induction methods,this paper proposes an induction method based on degree. Firstly the concept hierarchy relation is presented as a graph. Then the hierarchies of concepts are exploited with a reverse pruning algorithm. At last,after adding and pruning the concept relation the concept hierarchy is generated. Compared with variable value of parameter,the best performance,which is better than that of other methods,is reached when based on VSM but not using LSI.
同期刊论文项目
同项目期刊论文
期刊信息
位置：