基于图分析的领域知识获取技术  

Domain knowledge acquisition on basis of graph analysis

作　　者：张德政[1] 阿孜古丽[1] 刘洁卉[1] 

机构地区：[1]北京科技大学信息工程学院,北京100083

出　　处：《北京科技大学学报》2007年第S2期194-198,共5页Journal of University of Science and Technology Beijing

基　　金：国家"十一五"科技支撑计划基金资助项目(No2007BAI10B06);国家"863"计划基金资助项目(No2003AA115220)

摘　　要：知识获取技术制约着知识系统的研究和应用,有效地从文本中提取领域知识成为知识获取的重要途径.本文提出了基于本体和图分析的领域知识获取技术,分析了本体数据结构、本体概念的实例化以及基于图分析的语义场构造方法.建立了具有广泛适用性的文本知识获取系统框架,实现了原型系统.通过针对中医医案的中医领域知识获取实验验证,取得了较好的效果.Knowledge acquisition technique gives constrains to study and application of knowledge system,therefore,effective extraction of domain knowledge from text becomes the major approach to knowledge acquisition.Domain knowledge acquisition technique based on theories of ontology and graph analysis is proposed,data structure of ontology,instantiation of ontology concept and construction of semantic field are discussed.Frame of text knowledge acquisition system is constructed,which is of general applicability,and the prototype system has been realized.By experimental verification of domain knowledge acquisition in Traditional Chinese Medicine,the system achieved good results.

关 键 词：本体 图分析 知识获取 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]