关键词: 地理本体层次模型, 语义服务, 半自动语义标注, 语义相似性度量, 服务匹配与发现

Abstract:

As the applications of geographic information services and a widespread of users expand, the demand for geographic information services becomes increasingly diverse. On one hand, the amount of registered and released geographic information services on the internet rises rapidly; on the other hand, people are confused about how to efficiently discover and combine interested services to satisfy their demands. However, benefiting from the convenience of geospatial resource search through keywords and combining spatial filtering conditions of WCS (Web Catalog Service), users still frequently encounter problems of low comprehension and precision. Facing an increasingly large group of services, how to find the target service automatically, quickly and accurately has become the bottleneck for further development and application of geographic information service. In order to solve this problem in geographical information service matching and discovery, the paper adopted the information retrieval technology, semantic web technology and semantic service technology to achieve a good approach. As a whole, four aspects are discussed in depth, including the semantic annotation algorithm based on the geographical concept matching, the strategy of service descriptions for SOA framework, the geospatial semantic similarity measurement model, and the integrated algorism of the geospatial semantic similarity measurement and the service semantic matching that based on subsumption reasoning. The article firstly introduces the geographic concept annotation of geo-ontology hierarchical model, and then introduces a new semi-automatic semantic annotation algorithm that is based on geographic concept matching and increases the annotation efficiency. Considering the application requirement of geographic information service discovery, this research illustrates a new geographic concept similarity measurement model according to the description logic, which sets up a foundation to solve the semantic service problem regarding to similarity measurement. By compromising the semantic service similarity measurement model and traditional subsumption reasoning engine, this research develops a new geographic information service discovery and matching method, which will increase the discovery and matching efficiency evidently. In the end, the article presents the necessity to build an evaluation index system to evaluate the resultant quality on both geographic information service semantic annotation and service discovery and matching. And an application framework for geographic information service is introduced, which integrates geographic information semantic service annotation, service registry, catalog management, and service matching and discovery all together. Through analysis of practical cases, the article provides useful new research ideas for improving current geographic information service matching methods.

Key words: geo-ontology hierarchical model, semantic web service, semi-automatic semantic annotation, semantic similarity measurement, service matching and discovery