CAJ | 학술논문

本体在知识管理和语义网中扮演着越来越重要的角色，但本体的构建是一项费力又费时的工作，为此本文提出一种基于文本分类模式下的本体自动构建方法。该方法以形式化概念理论分析作为基础，计算出概念之间的相关度，以概念间的相关度和它们在文档中出现的频率矩阵构建本体概念图。实验结果表明，在文档分类模式下自动构建本体支持目前的信息分类系统，形成的本体有利于更好地共享和重用，促进语义Web的本体的升级。
본체재지식관리화어의망중분연착월래월중요적각색，단본체적구건시일항비력우비시적공작，위차본문제출일충기우문본분류모식하적본체자동구건방법。해방법이형식화개념이론분석작위기출，계산출개념지간적상관도，이개념간적상관도화타문재문당중출현적빈솔구진구건본체개념도。실험결과표명，재문당분류모식하자동구건본체지지목전적신식분류계통，형성적본체유리우경호지공향화중용，촉진어의Web적본체적승급。
Ontology is playing an increasingly important role in knowledge management and the semantic Web. However, creating ontologies is usually tedious and costly from scratch, this study presented a novel approach of automated ontology constructing for text information classification. This methodology used a degree of relevancy among concept that calculated through formal concept analysis and frequency matrix of terms appearing in document to structure an ontology diagram. An experiment about this research that based on the present information classification system was effective to automated ontology constructing, the ontologies accelerated information shared and reused to promote upgrading of semantic web.