作　　者： ; ; ; ;

机构地区： 深圳大学信息工程学院

出　　处： 《中山大学学报（自然科学版）》 2005年第3期15-19,共5页

摘　　要： 提出建立本体的协作-挖掘方法,即领域专家、知识工程师、领域终端用户通过互联网,协作建立本体。利用网络爬虫从语义网搜索并收集RDF文档或片段,自动分析初步建立RDFDB。自动分析领域专家、领域终端用户使用半形式化语言RDFL书写的实例文档和它们使用引导程序输入的记录,完善RDFDB。检验清洗RDFDB数据,并设计本体挖掘算法挖掘产生初始领域本体。挖掘算法使用红黑树建立索引,最坏时间复杂度为O(n3log2n)。验证、评估初始本体产生领域本体,并产生文档说明。建立新本体时,可合并RDFDB集成现有本体。在实验系统中,应用该方法建立计算机硬件信息领域本体。实验结果表明该方法是可行和高效的。 A collaborative_mining approach to building ontology is proposed.DEs (Domain Experts), KEs (Knowledge Engineers) and DEUs(Domain End Users) work on internet corporately in order to build ontology. Documents/fragments of RDF collected by Crawler serve as primitive resources. Instance documents written by DEs/DEUs in Resource Description Framework Language, which is a pseudo_formal language to describe instances, act as complementary resources. And records input using Wizard are optional resources. All these resources are analyzed automatically to establish the RDFDB. Data in RDFDB are refined before primitive domain ontology is mined. The worst time complexity of ontology mining algorithm is O(n^3log^2n), because its index service is based on Red_Black tree. Primitive ontology evolves into domain ontology after verification and evaluation.Explanatory documents are generated automatically. When building new ontology, RDFDB of existing ontologies can be integrated into RDFDB of the new one. The approach is adopted in building computer hardware ontology. The experiment results show that this approach is feasible and efficient.

关 键 词： 本体建立 协作 挖掘

领　　域： [自动化与计算机技术] [自动化与计算机技术]