摘要
现有地址输入提示方法涉及标准地址和POI的研究较少,地址字符串的索引大多采用Trie(字典)树索引,但Trie树建立时内存消耗巨大,在面临海量数据时,问题更加突出。针对以上问题,提出一种基于key-value数据库的快速地名地址输入提示方法。该方法基于Trie树结构进行改进,降低了地址索引的复杂度;基于keyvalue数据库构建Trie树,避免了内存消耗巨大的问题。实验结果表明,基于key-value数据库构建的Trie树索引较基于内存构建的Trie树索引在事务响应性能和内存消耗方面具有明显的优势和效率。
The existed address searching pre-prompt for standard address and POI information is relatively few. For the index of the address string, most will use Trie tree to index address. But Trie tree consumes huge memory in its establishment, especially when facing with massive data, the problem is particularly prominent. In view of these problems, this paper proposed a rapid address searching pre-prompt method based on key-value database. The method improved the structure of Trie tree, and reduced the complexity of address index. At the same time, the method avoided consuming huge memory in Trie tree establishment because of basing on key-value database to build the Trie tree. The experiment results show that the Trie tree index based on the key-value database has obvious advantages and efficiency in terms of transaction response performance and memory consumption compared with the Trie tree index based on memory construction.
出处
《计算机应用研究》
CSCD
北大核心
2017年第11期3334-3338,3344,共6页
Application Research of Computers
基金
警用地理信息技术公安部重点实验室开放基金资助项目(2016LPGIT05)
公安部重点研究计划课题(2013022DYJ018)
国家科技支撑计划资助项目(2012BAH35B03)
测绘遥感信息工程国家重点实验室专项科研经费资助项目