语义引导的图标式地理概念建模环境初探

Semantics guided geographic conceptual modeling environment based on icons

以地理问题求解和地理科学研究环境建设的实际需求为引导,针对当前地理建模过程中存在的建模思想难以交流与重用、建模方式复杂、多领域专家协同建模困难等问题,研究地理概念建模过程中地理概念场景、概念实体及其相互作用关系的表达与元数据描述方法,利用空间数据表达规范与地理模型元数据表达规范逐步引导数据与模型的选择与匹配,构建地理概念模型,并在此基础上提出一种可视化、引导式的面向地理问题表达的概念建模方法,为地理研究提供一个语义引导的图标式地理概念建模环境。

There are many problems in conceptual modeling process, such as difficulty of modeling conception sharing, complexity of modeling method, difficulty in cooperative modeling by experts from different domains, and so on. In order to solve those problems, we study the metadata and semantic of geographic conceptual scenarios, geographic conceptual entities and relationships among them, In this paper, geographic conceptual entities are represented by the configurable icons, the relationships among entities are represented while geographic conceptual scenario is organized, and finally the geographic conceptual scenario is built. As the manifestation of geographic conceptual entities and relationships among them, those icons contain the meaning of geo-model, geo-semantics, geodata, and the match rules between models and data. According to the geo-problem domain, geo-study area, and temporal and spatial scale, geographic icons are chosen to represent geographic conceptual entities and relationships, and match rules are used to build geographic conceptual scenarios. During this process, spatial data specification and geographic model metadata specification are designed as its induction and matching rules, and conception/relationship constrain engine, graphics/conception constrain engine and model/ data match engine are designed to guide the model match and data match process. In this paper, the geographic conceptual model of hydrological cycle at regional scale is applied to validate the geographic conceptual modeling environment. By the experiments, it is shown that our research is a new attempt for geographic conceptual modeling, and that this environment is flexible for geographic conceptual scenario building, modeling conception sharing, and modeling result reusing. As a result, it provides a convenient and visualized modeling method for experts in multi-domain of geographic fields.