基于 MIML 的 OGC 网络服务语义检索方法研究-《计算机技术与发展》

文章信息/Info

Title:: Research on Semantic Searching Method of OGC GeographicInformation Service Based on MIML

作者:: 孙君祯¹ ; 苗立志¹; 2* ; 徐兴永³; 1. 南京邮电大学地理与生物信息学院,江苏南京 210023;
2. 南京邮电大学江苏省智慧健康大数据分析与位置服务工程实验室,江苏南京 210023;
3. 南京邮电大学通信与信息工程学院,江苏南京 210003

Author(s):: SUN Jun-zhen¹ ; MIAO Li-zhi¹; 2* ; XU Xing-yong³; 1. School of Geographical and Biological Information,Nanjing University of Posts and Telecommunications, Nanjing 210023,China;
2. Jiangsu Engineering Laboratory for Smart Analysis of Healthy Big Data and Location Based Services, Nanjing University of Posts and Telecommunications,Nanjing 210023,China;
3. School of Telecommunications & Information Engineering,Nanjing University of Posts and Telecommunications, Nanjing 210003,China

Keywords:: multi-instance multi-label; OGC web service; semantic search; TF-IDF algorithm; geo-ontology

摘要:: 由于地理信息服务存在能力描述内容敏感度不够、语义信息缺乏等问题,地理信息服务的语义检索一直是空间数据搜索方面的研究热点。该文以 OGC( open geospatial consortium) 地理信息服务为研究对象,通过解析其能力文件中的相关标签的属性值, 采用 TF-IDF 算法提取能够描述 OGC 网络服务(OGC web service, OWS) 特征的标记词汇; 并基于MIMLBoost 算法实现对 OWS 的语义标记,设计了面向 OGC 地理信息服务的语义检索模型。结合 MIMLBoost 算法、TF-IDF算法和地质本体库, 设计并开发了面向 OGC 地理信息服务的语义检索系统, 实现了 OWS 数据的语义检索。通过对OneGeology 和美国地质勘探局( USGS)中 300 条可访问的地理信息服务数据进行检索实验,结果表明:基于 MIMLBoost 和TF-IDF 的 OGC 地理信息服务检索方法的查全率与查准率达到了 92% 、93. 48% ,与基于 GeoNetwork 的查询相比分别提高22% 和 16. 34% ,为面向 OWS 的语义检索方法的研究提供了一种解决方法。

Abstract:: The semantic retrieval of geographic information services has always been a hot research topic in spatial data search because of the problems such as insufficient sensitivity of capability file and lack of semantic information. Taking the OGC ( open geospatial consortium) geographic information service as the research object,the TF-IDF algorithm is used to extract the marked vocabularies that can describe the characteristics of the OGC Web Service ( OGS web service,OWS) by parsing the attribute values of the relevant tags in their capability file. And,the semantic labeling of OWS is realized by the MIMLBoost algorithm,which facilitates to develop a semantic retrieval model for OGC geographic information service. Combining MIMLBoost algorithm,TF-IDF algorithm and geological ontology base,a semantic retrieval system for OGC geographic information service is designed and developed,which realizes the semantic retrieval of OWS data. According to the retrieval experiment on 300 accessible geographic information service data from One Geology and USGS,the recall ratio and accuracy have reached 92% and 93. 48% based on the proposed retrieval method,increased by 22% and 16. 34% respectively compared with GeoNetwork-based queries, which provides a solution for the research of OWS - oriented semantic retrieval methods.