ElasticSearch安装IK分词插件以及自定义分词

港控/mmm° 2023-10-11 15:07 39阅读 0赞

### 一、ES安装IK分词插件 ###

*  基于docker容器的方式启动ES实例  
    启动脚本如下：

docker run -d -p9200:9200 -p9300:9300 -e "discovery.type=single-node" -v /opt/docker/es/plugins:/usr/share/elasticsearch/plugins/ -e ES_JAVA_OPTS="-Xms120m -Xmx120m" --name es01 elasticsearch:7.7.0

将容器的/usr/share/elasticsearch/plugins/目录挂载到虚拟机的/opt/docker/es/plugins目录。

*  根据ES的版本下载对应的IK分词器  
    [IK分词器github地址][IK_github]  
    解压后上传到虚拟机的/opt/docker/es/plugins目录

![在这里插入图片描述][20201128204424224.png]

chomd -R 777 ik #修改一些ik文件夹对应的权限

由于是具名挂载的，此时容器中的plugins目录也有了ik文件夹

*  重启ES实例

docker restart 容器ID

*  测试

1.  it\_max\_word测试  
    ![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70]
2.  ik\_smart  
    ![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70 1]

### 二、ES使用自定义词库 ###

有些网络流行新词，ES并不能准确识别例如下面的测试，如果是人为分词，最想要分为乔碧罗、电商、带货，这么三个词，此时就需要用自定义词库。  
![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70 2]

*  启动一个nginx实例，将/etc/nginx下的配置挂载到容器外部

docker container cp nginx:/etc/nginx .

*  关闭刚才启动的nginx，再次启动一个nginx实例，将配置挂载到刚才的目录

docker run -d -p 6000:80 --name nginx -v /opt/docker/nginx/html:/usr/share/nginx/html -v /opt/docker/nginx/logs:/var/log/nginx -v /opt/docker/nginx/conf:/etc/nginx nginx

*  进入/opt/docker/nginx/html目录，创建words目录，创建words.txt文件，输入要分词的内容

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70 3]

*  本机测试  
    ![在这里插入图片描述][2020112911565610.png]
 *  修改es外部挂载的plugins目录下的IK分词配置  
    ![在这里插入图片描述][20201129115825823.png]  
    ![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70 4]
 *  重启容器再次测试 `docker restart 容器ID`

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70 5]

[IK_github]: https://github.com/medcl/elasticsearch-analysis-ik/releases/
[20201128204424224.png]: https://img-blog.csdnimg.cn/20201128204424224.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70]: https://img-blog.csdnimg.cn/20201128204628772.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70 1]: https://img-blog.csdnimg.cn/2020112820472393.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70 2]: https://img-blog.csdnimg.cn/20201129111725132.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70 3]: https://img-blog.csdnimg.cn/20201129115554421.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2,size_16,color_FFFFFF,t_70
[2020112911565610.png]: https://img-blog.csdnimg.cn/2020112911565610.png
[20201129115825823.png]: https://img-blog.csdnimg.cn/20201129115825823.png
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70 4]: https://img-blog.csdnimg.cn/20201129115850769.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2_size_16_color_FFFFFF_t_70 5]: https://img-blog.csdnimg.cn/20201129112655917.png?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L3FxXzQzNzUwNjU2,size_16,color_FFFFFF,t_70