MySQL 子串模糊搜索

【问题标题】：MySQL SubString Fuzzy SearchMySQL 子串模糊搜索
【发布时间】：2017-01-01 22:09:09
【问题描述】：

我有一个非常有趣的问题：

我有一个 MySQL 表“Venue”，其中包含以下字段：“name”、“addressLine1”、“addressLine2”、“addressLine3”、“city”、“country”、“description”；所有字段都是 VARCHAR。 'description' 是一个较大的文本字段。

我想做的是在餐桌上进行模糊搜索。到目前为止，我正在使用：

SELECT * FROM Venue WHERE MATCH(name, addressLine1,..., description) AGAINST("London" IN NATURAL LANGUAGE MODE).

我还可以根据 MA​​TCH 分数对该查询进行排序。

这很好，但有一些明显的问题： 1) 如果用户键入“lond”，则不返回任何内容 2) 如果用户键入“lodnod”，则不返回任何内容。

我想到的另一种解决方案是使用 Levenshtein 扩展（2 个动作） - 因此对于搜索词“lodnod”，查询将类似于：

SELECT * FROM Venue WHERE Venue. name LIKE '%__lodnod%'
OR Venue.addressLine1 LIKE '%__lodnod%'
OR Venue.addressLine2 LIKE '%__lodnod%'
OR Venue.addressLine3 LIKE '%__lodnod%'
OR Venue.city LIKE '%__lodnod%'
OR Venue.county LIKE '%__lodnod%'
OR Venue.country LIKE '%__lodnod%'
OR Venue. name LIKE '%lodnod%'
OR Venue.addressLine1 LIKE '%lodnod%'
OR Venue.addressLine2 LIKE '%lodnod%'
OR Venue.addressLine3 LIKE '%lodnod%'
OR Venue.city LIKE '%lodnod%'
OR Venue.county LIKE '%lodnod%'
OR Venue.country LIKE '%lodnod%'
OR Venue. name LIKE '%_lodnod%'
OR Venue.addressLine1 LIKE '%_lodnod%'
OR Venue.addressLine2 LIKE '%_lodnod%'
OR Venue.addressLine3 LIKE '%_lodnod%'
OR Venue.city LIKE '%_lodnod%'
OR Venue.county LIKE '%_lodnod%'
OR Venue.country LIKE '%_lodnod%'
OR Venue. name LIKE '%_odnod%'
OR Venue.addressLine1 LIKE '%_odnod%'
OR Venue.addressLine2 LIKE '%_odnod%'
OR Venue.addressLine3 LIKE '%_odnod%'
OR Venue.city LIKE '%_odnod%'
OR Venue.county LIKE '%_odnod%'
OR Venue.country LIKE '%_odnod%'
OR Venue. name LIKE '%__odnod%'
OR Venue.addressLine1 LIKE '%__odnod%'
OR Venue.addressLine2 LIKE '%__odnod%'
OR Venue.addressLine3 LIKE '%__odnod%'
OR Venue.city LIKE '%__odnod%'
OR Venue.county LIKE '%__odnod%'
OR Venue.country LIKE '%__odnod%'
OR Venue. name LIKE '%_l_odnod%'
OR Venue.addressLine1 LIKE '%_l_odnod%'
OR Venue.addressLine2 LIKE '%_l_odnod%'
OR Venue.addressLine3 LIKE '%_l_odnod%'
OR Venue.city LIKE '%_l_odnod%'
OR Venue.county LIKE '%_l_odnod%'
OR Venue.country LIKE '%_l_odnod%'
OR Venue. name LIKE '%_ldnod%'
OR Venue.addressLine1 LIKE '%_ldnod%'
OR Venue.addressLine2 LIKE '%_ldnod%'
OR Venue.addressLine3 LIKE '%_ldnod%'
OR Venue.city LIKE '%_ldnod%'
OR Venue.county LIKE '%_ldnod%'
OR Venue.country LIKE '%_ldnod%'
OR Venue. name LIKE '%_l_dnod%'
OR Venue.addressLine1 LIKE '%_l_dnod%'
OR Venue.addressLine2 LIKE '%_l_dnod%'
OR Venue.addressLine3 LIKE '%_l_dnod%'
OR Venue.city LIKE '%_l_dnod%'
OR Venue.county LIKE '%_l_dnod%'
OR Venue.country LIKE '%_l_dnod%'
OR Venue. name LIKE '%_lo_dnod%'
OR Venue.addressLine1 LIKE '%_lo_dnod%'
OR Venue.addressLine2 LIKE '%_lo_dnod%'
OR Venue.addressLine3 LIKE '%_lo_dnod%'
OR Venue.city LIKE '%_lo_dnod%'
OR Venue.county LIKE '%_lo_dnod%'
OR Venue.country LIKE '%_lo_dnod%'
OR Venue. name LIKE '%_lonod%'
OR Venue.addressLine1 LIKE '%_lonod%'
OR Venue.addressLine2 LIKE '%_lonod%'
OR Venue.addressLine3 LIKE '%_lonod%'
OR Venue.city LIKE '%_lonod%'
OR Venue.county LIKE '%_lonod%'
OR Venue.country LIKE '%_lonod%'
OR Venue. name LIKE '%_lo_nod%'
OR Venue.addressLine1 LIKE '%_lo_nod%'
OR Venue.addressLine2 LIKE '%_lo_nod%'
OR Venue.addressLine3 LIKE '%_lo_nod%'
OR Venue.city LIKE '%_lo_nod%'
OR Venue.county LIKE '%_lo_nod%'
OR Venue.country LIKE '%_lo_nod%'
OR Venue. name LIKE '%_lod_nod%'
OR Venue.addressLine1 LIKE '%_lod_nod%'
OR Venue.addressLine2 LIKE '%_lod_nod%'
OR Venue.addressLine3 LIKE '%_lod_nod%'
OR Venue.city LIKE '%_lod_nod%'
OR Venue.county LIKE '%_lod_nod%'
OR Venue.country LIKE '%_lod_nod%'
OR Venue. name LIKE '%_lodod%'
OR Venue.addressLine1 LIKE '%_lodod%'
OR Venue.addressLine2 LIKE '%_lodod%'
OR Venue.addressLine3 LIKE '%_lodod%'
OR Venue.city LIKE '%_lodod%'
OR Venue.county LIKE '%_lodod%'
OR Venue.country LIKE '%_lodod%'
OR Venue. name LIKE '%_lod_od%'
OR Venue.addressLine1 LIKE '%_lod_od%'
OR Venue.addressLine2 LIKE '%_lod_od%'
OR Venue.addressLine3 LIKE '%_lod_od%'
OR Venue.city LIKE '%_lod_od%'
OR Venue.county LIKE '%_lod_od%'
OR Venue.country LIKE '%_lod_od%'
OR Venue. name LIKE '%_lodn_od%'
OR Venue.addressLine1 LIKE '%_lodn_od%'
OR Venue.addressLine2 LIKE '%_lodn_od%'
OR Venue.addressLine3 LIKE '%_lodn_od%'
OR Venue.city LIKE '%_lodn_od%'
OR Venue.county LIKE '%_lodn_od%'
OR Venue.country LIKE '%_lodn_od%'
OR Venue. name LIKE '%_lodnd%'
OR Venue.addressLine1 LIKE '%_lodnd%'
OR Venue.addressLine2 LIKE '%_lodnd%'
OR Venue.addressLine3 LIKE '%_lodnd%'
OR Venue.city LIKE '%_lodnd%'
OR Venue.county LIKE '%_lodnd%'
OR Venue.country LIKE '%_lodnd%'
OR Venue. name LIKE '%_lodn_d%'
OR Venue.addressLine1 LIKE '%_lodn_d%'
OR Venue.addressLine2 LIKE '%_lodn_d%'
OR Venue.addressLine3 LIKE '%_lodn_d%'
OR Venue.city LIKE '%_lodn_d%'
OR Venue.county LIKE '%_lodn_d%'
OR Venue.country LIKE '%_lodn_d%'
OR Venue. name LIKE '%_lodno_d%'
OR Venue.addressLine1 LIKE '%_lodno_d%'
OR Venue.addressLine2 LIKE '%_lodno_d%'
OR Venue.addressLine3 LIKE '%_lodno_d%'
OR Venue.city LIKE '%_lodno_d%'
OR Venue.county LIKE '%_lodno_d%'
OR Venue.country LIKE '%_lodno_d%'
OR Venue. name LIKE '%_lodno%'
OR Venue.addressLine1 LIKE '%_lodno%'
OR Venue.addressLine2 LIKE '%_lodno%'
OR Venue.addressLine3 LIKE '%_lodno%'
OR Venue.city LIKE '%_lodno%'
OR Venue.county LIKE '%_lodno%'
OR Venue.country LIKE '%_lodno%'
OR Venue. name LIKE '%_lodno_%'
OR Venue.addressLine1 LIKE '%_lodno_%'
OR Venue.addressLine2 LIKE '%_lodno_%'
OR Venue.addressLine3 LIKE '%_lodno_%'
OR Venue.city LIKE '%_lodno_%'
OR Venue.county LIKE '%_lodno_%'
OR Venue.country LIKE '%_lodno_%'
OR Venue. name LIKE '%_lodnod_%'
OR Venue.addressLine1 LIKE '%_lodnod_%'
OR Venue.addressLine2 LIKE '%_lodnod_%'
OR Venue.addressLine3 LIKE '%_lodnod_%'
OR Venue.city LIKE '%_lodnod_%'
OR Venue.county LIKE '%_lodnod_%'
OR Venue.country LIKE '%_lodnod_%'
OR Venue. name LIKE '%dnod%'
OR Venue.addressLine1 LIKE '%dnod%'
OR Venue.addressLine2 LIKE '%dnod%'
OR Venue.addressLine3 LIKE '%dnod%'
OR Venue.city LIKE '%dnod%'
OR Venue.county LIKE '%dnod%'
OR Venue.country LIKE '%dnod%'
OR Venue. name LIKE '%_dnod%'
OR Venue.addressLine1 LIKE '%_dnod%'
OR Venue.addressLine2 LIKE '%_dnod%'
OR Venue.addressLine3 LIKE '%_dnod%'
OR Venue.city LIKE '%_dnod%'
OR Venue.county LIKE '%_dnod%'
OR Venue.country LIKE '%_dnod%'
OR Venue. name LIKE '%o_dnod%'
OR Venue.addressLine1 LIKE '%o_dnod%'
OR Venue.addressLine2 LIKE '%o_dnod%'
OR Venue.addressLine3 LIKE '%o_dnod%'
OR Venue.city LIKE '%o_dnod%'
OR Venue.county LIKE '%o_dnod%'
OR Venue.country LIKE '%o_dnod%'
OR Venue. name LIKE '%onod%'
OR Venue.addressLine1 LIKE '%onod%'
OR Venue.addressLine2 LIKE '%onod%'
OR Venue.addressLine3 LIKE '%onod%'
OR Venue.city LIKE '%onod%'
OR Venue.county LIKE '%onod%'
OR Venue.country LIKE '%onod%'
OR Venue. name LIKE '%o_nod%'
OR Venue.addressLine1 LIKE '%o_nod%'
OR Venue.addressLine2 LIKE '%o_nod%'
OR Venue.addressLine3 LIKE '%o_nod%'
OR Venue.city LIKE '%o_nod%'
OR Venue.county LIKE '%o_nod%'
OR Venue.country LIKE '%o_nod%'
OR Venue. name LIKE '%od_nod%'
OR Venue.addressLine1 LIKE '%od_nod%'
OR Venue.addressLine2 LIKE '%od_nod%'
OR Venue.addressLine3 LIKE '%od_nod%'
OR Venue.city LIKE '%od_nod%'
OR Venue.county LIKE '%od_nod%'
OR Venue.country LIKE '%od_nod%'
OR Venue. name LIKE '%odod%'
OR Venue.addressLine1 LIKE '%odod%'
OR Venue.addressLine2 LIKE '%odod%'
OR Venue.addressLine3 LIKE '%odod%'
OR Venue.city LIKE '%odod%'
OR Venue.county LIKE '%odod%'
OR Venue.country LIKE '%odod%'
OR Venue. name LIKE '%od_od%'
OR Venue.addressLine1 LIKE '%od_od%'
OR Venue.addressLine2 LIKE '%od_od%'
OR Venue.addressLine3 LIKE '%od_od%'
OR Venue.city LIKE '%od_od%'
OR Venue.county LIKE '%od_od%'
OR Venue.country LIKE '%od_od%'
OR Venue. name LIKE '%odn_od%'
OR Venue.addressLine1 LIKE '%odn_od%'
OR Venue.addressLine2 LIKE '%odn_od%'
OR Venue.addressLine3 LIKE '%odn_od%'
OR Venue.city LIKE '%odn_od%'
OR Venue.county LIKE '%odn_od%'
OR Venue.country LIKE '%odn_od%'
OR Venue. name LIKE '%odnd%'
OR Venue.addressLine1 LIKE '%odnd%'
OR Venue.addressLine2 LIKE '%odnd%'
OR Venue.addressLine3 LIKE '%odnd%'
OR Venue.city LIKE '%odnd%'
OR Venue.county LIKE '%odnd%'
OR Venue.country LIKE '%odnd%'
OR Venue. name LIKE '%odn_d%'
OR Venue.addressLine1 LIKE '%odn_d%'
OR Venue.addressLine2 LIKE '%odn_d%'
...(cut short because of maximum 30000 character limit)
OR Venue.city LIKE '%lodno_%'
OR Venue.county LIKE '%lodno_%'
OR Venue.country LIKE '%lodno_%'
OR Venue. name LIKE '%lodno__%'
OR Venue.addressLine1 LIKE '%lodno__%'
OR Venue.addressLine2 LIKE '%lodno__%'
OR Venue.addressLine3 LIKE '%lodno__%'
OR Venue.city LIKE '%lodno__%'
OR Venue.county LIKE '%lodno__%'
OR Venue.country LIKE '%lodno__%'
OR Venue. name LIKE '%lodno_d_%'
OR Venue.addressLine1 LIKE '%lodno_d_%'
OR Venue.addressLine2 LIKE '%lodno_d_%'
OR Venue.addressLine3 LIKE '%lodno_d_%'
OR Venue.city LIKE '%lodno_d_%'
OR Venue.county LIKE '%lodno_d_%'
OR Venue.country LIKE '%lodno_d_%'
OR Venue. name LIKE '%lodno%'
OR Venue.addressLine1 LIKE '%lodno%'
OR Venue.addressLine2 LIKE '%lodno%'
OR Venue.addressLine3 LIKE '%lodno%'
OR Venue.city LIKE '%lodno%'
OR Venue.county LIKE '%lodno%'
OR Venue.country LIKE '%lodno%'
OR Venue. name LIKE '%lodnod__%'
OR Venue.addressLine1 LIKE '%lodnod__%'
OR Venue.addressLine2 LIKE '%lodnod__%'
OR Venue.addressLine3 LIKE '%lodnod__%'
OR Venue.city LIKE '%lodnod__%'
OR Venue.county LIKE '%lodnod__%'
OR Venue.country LIKE '%lodnod__%'
OR Venue. name LIKE '%lodnod_%'
OR Venue.addressLine1 LIKE '%lodnod_%'
OR Venue.addressLine2 LIKE '%lodnod_%'
OR Venue.addressLine3 LIKE '%lodnod_%'
OR Venue.city LIKE '%lodnod_%'
OR Venue.county LIKE '%lodnod_%'
OR Venue.country LIKE '%lodnod_%'

显然，这个查询是 1) 将对服务器造成巨大负载，响应时间将受到影响。 2) 我不知道如何评分并因此对其进行排序。

有没有办法更好地做到这一点？

【问题讨论】：

    标签： mysql fuzzy-search


    【解决方案1】：

    MySQL's documentation page for full text search functions 有一些有用的信息。尽管上面的 cmets 说了什么，您不需要将自己限制在 MySQL 中的 MyISAM 表类型。正如其他人所说，MySQL 可能不是最佳选择，但您可以采取一些措施使其更实用。

    列聚合

    如果您发现自己总是对多个列进行这样的搜索，那么拥有一个仅包含来自所有其他列的标记的“可搜索”列可能会有所帮助。您需要一些机制来使其保持最新状态，但ON INSERT 和ON UPDATE 触发器就足够了。这可能会简化您的索引并使您的查询更易于阅读（并有望加快响应速度），但并不能解决您最初的问题。

    这种“可搜索”列理念的一个优点是您可以悄悄地添加常见拼写错误的标记，以使您的用户更轻松。您的搜索日志在这里很有帮助。

    考虑使用 Soundex

    最近取得了一些成功的解决方案是使用 soundex 而不是 Levenshtein 扩展。参见，例如，Rob Gravelle (2015) "MySQL Fuzzy Text Searching Using the SOUNDEX Function，数据库日志

    用户定义函数 (UDF)

    另一个选项是在您的 MySQL 实现中添加用户定义函数 (UDF)。参见例如this Stack Overflow question 或this blog post。

    无论您如何接近它，您都会发现使用 Levenshtein 进行真正的模糊搜索是一种缓慢的野兽。祝你好运！

    【讨论】：

      【解决方案2】：

      不要使用 mysql 来处理这些东西。您需要一个为索引搜索材料优化的引擎。为此，您可以使用elasticsearch 或solr 并使用mysql 对其进行索引。 Mysql 根本就不是为了这个目的而设计的

      【讨论】：

      • 感谢您的回复 - 我知道这是我可能必须采取的路线 - 但假设我打算在 MySQL 中这样做 - 有没有办法？？？
      • 如果你使用 MyIsam 表类型是可能的。它具有全文搜索索引。但正如我提到的，它远非最佳dev.mysql.com/doc/refman/5.7/en/myisam-storage-engine.html 供参考
      • 是的 - 全文搜索是我现在使用的，但它真的不是我要找的。​​span>
      猜你喜欢
      • 1970-01-01
      • 1970-01-01
      • 2016-07-01
      • 1970-01-01
      • 2011-10-30
      • 2020-03-15
      • 2014-08-20
      • 2011-10-28
      • 2016-06-12
      相关资源
      最近更新 更多
      热门标签
      Java Python linux javascript Mysql C# Docker 算法 前端 SpringBoot Redis Vue spring 设计模式 .net core .net kubernetes c++ 数据库 数据结构 大数据 js 机器学习 微服务 Android Go 程序员 面试 JVM ASP.net core 云原生 人工智能 后端 PHP git CSS golang k8s Nginx Django mybatis 深度学习 多线程 React 架构 devops 爬虫 云计算 Spring Boot LeetCode