Abstract: The authors present a new approach for modeling qualitative geographic information retrieval based on the qualitative representation, semantic matching and qualitative reasoning methods. Information in web document and user queries are represented and extracted by a new qualitative method considering the thematic and geographic semantics synthetically. The novel retrieval method is used to retrieve the qualitative geographic information to support the semantic matching and results ranking. Finally, a prototype system is implemented to evaluate this research. Through these works, the geographic information retrieval technology is much adaptive to human commonsense cognition and behaviors. The efficiency of geographic information retrieval is improved with the qualitative representation and qualitative spatial reasoning method.

Key words: geographic information retrieval, qualitative model, geographic information system

摘要： 针对当前定量化的地理信息检索模型无法有效处理自然语义导致检索结果不理想的问题, 以语义匹配为原则, 以定性表达为基础, 以推理方法为手段, 提出基于定性空间推理的定性地理信息检索的方法及其形式化模型, 实现Web文档信息内容与查询请求的定性表达和信息提取, 支持基于语义的检索匹配和结果排序, 并进行原型系统的实验验证。这种定性检索方法符合人类常识性认知和表述习惯的方式, 基于定性知识表达和推理决策技术的地理信息检索问题解决方法可以提高现代地理信息检索技术的有效性。

关键词: 地理信息检索, 定性表达, 地理信息系统