地质年代本体及其在语义检索中的应用

Geologic Time Scale Ontology and Its Applications in Semantic Retrieval

不同国家、语言和版本的地质年代概念体系中普遍存在的语义异构问题,以及概念之间、概念与数据之间缺少语义关联等问题,阻碍了概念的准确理解和应用,难以满足数据集成、检索的应用需求。地质年代本体是解决这些问题的有效方式。但现有研究主要针对国际地质年代概念体系,其内容和表达方式不完全适用于国内的概念体系。本文提出以中国地质年代和地层概念体系为主的地质年代本体,设计了本体中各概念属性信息,尤其是时间信息的语义表达方法。在此基础上采用模块化方法相应构建了本体,并设计实现了一个本体应用原型系统,提供了以地质年代本体知识库为基础的概念查询检索功能,以及融合全文检索技术和地质年代本体的数据语义检索功能。应用案例的初步结果表明,本研究构建的地质年代本体能为用户提供表现形式多样的知识查询服务,能较好地解决地学概念与数据中存在的语义异构、关联缺失问题,更为智能、完整、准确地完成地学数据的语义检索。

Different countries or organizations publish different versions or language editions of geologic time scales. Problems of ambiguity, such as synonymy or near-synonym and polysemy or other kinds of semantic heterogeneity arise in those geochronological concept systems, along with the lack of semantic linkage among different concepts and data, have hindered people from understanding and using those concepts accurately. Also, this caused insufficient results of integration and retrieval of geo-data for users＇ requirements. The ontology of geologic time proved to be an effective solution to these problems. However, current studies focus on international geologic time scales, and their content and formalizations are not fully applicable in China. In this paper, we present a new Geologic Time Scale Ontology which mainly focuses on Chinese geochronological and stratigraphic concept systems. It describes the formalizations of attributes, especially the temporal features and relations of those concepts. It adopts a modular method to build this ontology. In addition, we propose a design of a pilot system to study the utility of this ontology as the basis of a geoscience knowledge graph in knowledge retrieval. Furthermore, we implemented an approach of semantic geo-data retrieval in the pilot system which uses a hybrid strategy of fulltext and ontology-based search. Applications of knowledge graph and geo-data retrieval based on the abilities such as temporal reasoning of Geologic Time Scale Ontology, proved that theproposed researches are effective in resolving semantic heterogeneity issues in geoscience knowledge and data. They can not only facilitate discovery of geoscience knowledge but also achieve the function of semantic data retrieval more intelligently, comprehensively and accurately for the discovery of specified and relevant data.