用户名: 密码: 验证码:
基于领域本体的地质信息元数据语义挖掘与知识发现研究
详细信息    查看官网全文
摘要
随着地质信息技术的逐渐成熟,地质信息元数据的网络发布成为了地质信息服务的有效方式。为了解决面向用户提供智能检索和面向机构提供机器可读可理解的语义关联服务等问题,基于领域本体的地质信息核心元数据的语义数据挖掘和知识发现是必要的。本课题的主要研究内容有三个方面:首先,提出了基于叙词表构建领域本体的半自动化方法。其通过对非规范知识的词构解析及无监督分类,实现了非规范知识的领域本体构建问题,解决了非规范知识的智能检索的关键问题。其次,提出了基于领域本体构建含权值的树进行计算两两资料关键词向量的相关度,对检索结果进行排序,以提供更高的检索效率。最后,基于D2R实现了核心元数据的关联数据发布,面向机构提供机器可读可理解的语义关联服务。本研究将智能检索和语义网Web服务应用到地质信息元数据服务领域中,不仅将与用户愿望最相关的文献排到最前面,提升了用户的体验,更实现了语义网Web服务。
With the maturing of the geological information technology, the online publication of geological information metadata has become an effective way for the geological information service. To solve the existing problems like user-oriented intelligent retrieval supply and institution-oriented correlation service, it is highly necessary to conduct research on semantic data mining and knowledge discovery of geological information metadata based on domain ontology. This project includes three main aspects: firstly, it aims to propose a semi-automatic method based on the thesaurus construction of domain ontology. Through the structure parsing and unsupervised classification of the non-canonical knowledge, the domain ontology of the non-canonical knowledge will be constructed and the intelligent retrieval of the non-canonical knowledge will be solved. Secondly, we aim to build a model to calculate the correlation between any two documents and form a correlation vector, which is based on a domain ontology tree with different node value. The final correlation is sorted and stored to provide higher search efficiency. Finally, based on D2 R technology, the linked data of core metadata will be released, and the semantics associated service that is machine-readable and machine-understandable will be provided. In this study overall, an intelligent search and semantic web service will be applied to geological information metadata services field, which will not only route the user's most wanted documents to the front, but also will allow the semantic web services.
引文

© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号

地址:北京市海淀区学院路29号 邮编:100083

电话:办公室:(+86 10)66554848;文献借阅、咨询服务、科技查新:66554700