ElasticSearch分布式搜索引擎在地名地址检索中的应用

Application of ElasticSearch Distributed Search Engine in Place Name Address Retrieval

针对传统全文搜索引擎Lucene无法满足海量地名地址数据高速分布式检索的问题,本文通过对分布式搜索引擎ElasticSearch的集群架构、索引分片与复制集、地理要素索引存储、数据检索及位置聚合等关键点进行详细研究,结合地名地址库数据结构的特点,提出了一种面向Web电子地图开发的分布式地名地址搜索服务构架,实现了地名地址模糊搜索、专题搜索、视野内搜索、周边搜索等功能。该架构具备可伸缩性、容错性、支持大规模并发访问的特点。

According to the fact that the traditional full-text search engine Lucene could not satisfy the high-speed distributed retrieval of massive address data,combined with the data structure features of the geographical address database,a distributed geographical address search service architecture for Web electric map developing was proposed and implemented in this paper through further studying the distributed clusters of ElasticSearch,index sharing,geolocation index storage and aggregations.The system has the following functions:fuzzy searching,specialized searching,intra-view searching,and surrounding searching.This architecture is scalable,fault-tolerant,and supports large-scale concurrent access.