地址编码是指将地址映射成地理坐标的过程, 它提供了一种把描述性地理位置信息转换成GIS 地理坐标的方式。本文分析了当前标准规范中的地址编码规则, 认为街道与小区等院落实体作为同级别的区域限定物弱化了街道的空间特征, 并且在现有地址编码规则基础上提出了将街道前置作为区域限定物的地址编码改进方案;进一步, 为了体现街道作为各类地理实体的空间参考基础, 本文按照街道的前进方向, 定义了在街道线性空间中的地理实体链表组织模型, 实现了将相互独立的地址编码实体在街道中的串联组织, 从而为今后地址查询中的参考定位提供了模型支持;最后, 给出了街道两侧不同类型地理实体纳入街道地址编码实体链表的实例说明。

Address geocoding is a kind of address matching method based on spatial technique. It is the process of transforming a description of a location, such as coordinates, the address or the name of a place, to a location on the earth’s surface. Address geocoding technologies started late in China, many cities just start capturing the geo coding address data. Many topics need to be explored, such as address code structure and geo coding data management. The authors analyzed the current address geocoding rules. The street is regarded as the same level distinct restriction as the courtyard in the current strardards. At this point they think that the spatial characters of the street are weakened. On the basis of relevant national standards, they put forward the definition amendment of city standard address geocoding. In the amendment, as distinct restriction, street becomes independent and is in front of courtyard. Thus, an entity must belong to one street and the address code must contain the street’s code. Usually it is difficult to do address matching with many types of address data because the address data store in different database tables by the entity types. Using streets as spatial reference system, according to the street onward direction, they well integrate the doorplates, address points and parcels together and put forward the new unified street geocoding model: street address chain model. In this model, geocoding is not simple address string matching any more, it also makes well use of the streets’direction, parcels’adjacency and other spatial characters to assist address geocoding process. This can provide model support for future address matching. The example about street address chain model of address geocoding is given. Predictably, this new address geocoding model will improve the match rate and accuracy greatly.