【摘要】 在研究分析地址模型的基础上,建立了存储标准地址数据集的标准地址库和自定义的地址匹配规则库,提出了一种基于规则的模糊中文地址编码方法。该方法在依据标准地址库分词的同时,也沿着自定义的地址匹配规则进行推理,从而缩小了下次分词所用到的目标数据集,提高了系统执行效率。另外,通过借助构建的规则树与歧义栈,提高了文中定义的两类模糊地址匹配的成功率。最后,基于该算法建立了一个地理编码原型系统,并利用经济普查项目中的相关数据对算法的可用性进行了验证。更多还原

【Abstract】 After analyzing Chinese address model,this paper built a standard address database and an address matching rules database,and then presented a rule-based Geocoding method for fuzzy Chinese addresses.This method used the standard address database to segment the input fuzzy Chinese address.At the same time,the method used the rules database to reduce and find a standard address that matched with that fuzzy address.The method used the customized rules to reduce candidate addresses so that it can participate in match reduction and save the matching executive time.In addition,the introduction of rule tree and semantic stacks also promote the matching of fuzzy address.Finally,a Geocoding prototype system was built,and then its availability was verified utilizing the data of natural economic census project.更多还原