面向更新信息提取与集成的空间实体匹配方法

A spatial entity matching approach for updates extraction and integration

本文为了自动发现提取新版基础地理数据层中的更新变化信息,并将其集成到用户数据层中,提出了一种依据属性、几何、拓扑关系等多源信息的空间实体复合式匹配方法。首先,阐述了该方法的基本过程以及确定实体是否匹配的判别规则;然后,在定义空间实体主要变化类型的基础上,讨论了根据匹配结果提取新增、消失等变化实体及其更新信息的基本策略;最后,介绍了在匹配结果指引下,通过插入、更新、删除等三个基本操作及其组合集成更新信息的基本方式。实验结果表明,所提方法及相关策略稳定可行,能够较全面地匹配两个时间跨度较大数据层之间的对应实体,有效弥补传统方法仅利用某一类信息,容易产生漏配的不足。

To automatically extract updates from the new version of fundamental geospatial data layer and integrate them into user data layer, a composite spatial entity approach based on multi-source information including attribute, geometry and topology relationship was proposed. Firstly, its basic process and the rules to decide whether entities were corresponding are discussed. Then, the strategies for extracting updates such as creation, disappearance according to the entity matching results were designed based on the definition of change types of spatial entities. Finally, the basic ways of integrating updates guided by the matching results was explained, which used three basic operations （insert, update and delete） and their combinations. The experimental results showed that the proposed ap- proach and related strategies were stable and feasible, and could match the corresponding entities between data layers with the longer time interval and overcome the limitations of the traditional approaches such as being prone to omit the real matches.