一、概要：
1.es默认的分词器对中文支持不好，会分割成一个个的汉字。ik分词器对中文的支持要好一些，主要由两种模式：ik_smart和ik_max_word
2.环境
操作系统：centos
es版本：6.0.0
二、安装插件
1.插件地址：https://github.com/medcl/elasticsearch-analysis-ik
2.运行命令行：
./bin/elasticsearch-plugin install https://github.com/medcl/elasticsearch-analysis-ik/releases/download/v6.0.0/elasticsearch-analysis-ik-6.0.0.zip
运行完成后会发现多了以下文件：esroot 下的plugins和config文件夹多了analysis-ik目录。
三、重启es
1.查找es进程
ps -ef | grep elastic
2.终止进程
从上面的结果可以看到es进程号是12776.
执行命令：
kill 12776
3.启动es后台运行
./bin/sh elastic search –d
提醒：重启es会重新分片，线上环境要注意了。
四、测试
1.使用ik_max_word分词
GET _analyze { "analyzer":"ik_max_word", "text":"*国歌" }
分词结果：
{ "tokens": [ { "token": "*", "start_offset": 0, "end_offset": 7, "type": "CN_WORD", "position": 0 }, { "token": "中华人民", "start_offset": 0, "end_offset": 4, "type": "CN_WORD", "position": 1 }, { "token": "中华", "start_offset": 0, "end_offset": 2, "type": "CN_WORD", "position": 2 }, { "token": "华人", "start_offset": 1, "end_offset": 3, "type": "CN_WORD", "position": 3 }, { "token": "人民*", "start_offset": 2, "end_offset": 7, "type": "CN_WORD", "position": 4 }, { "token": "人民", "start_offset": 2, "end_offset": 4, "type": "CN_WORD", "position": 5 }, { "token": "*", "start_offset": 4, "end_offset": 7, "type": "CN_WORD", "position": 6 }, { "token": "共和", "start_offset": 4, "end_offset": 6, "type": "CN_WORD", "position": 7 }, { "token": "国", "start_offset": 6, "end_offset": 7, "type": "CN_CHAR", "position": 8 }, { "token": "国歌", "start_offset": 7, "end_offset": 9, "type": "CN_WORD", "position": 9 } ] }
2.使用ik_smart分词
GET _analyze { "analyzer":"ik_smart", "text":"*国歌" }
分词结果：
{ "tokens": [ { "token": "*", "start_offset": 0, "end_offset": 7, "type": "CN_WORD", "position": 0 }, { "token": "国歌", "start_offset": 7, "end_offset": 9, "type": "CN_WORD", "position": 1 } ] }
五、java api分词测试
1.调用ik_max_word分词
@Test public void analyzer_ik_max_word() throws Exception { java.lang.String text = "提前祝大家春节快乐！"; TransportClient client = EsClient.get(); AnalyzeRequest request = (new AnalyzeRequest()).analyzer("ik_max_word").text(text); List<AnalyzeResponse.AnalyzeToken> tokens = client.admin().indices().analyze(request).actionGet().getTokens(); System.out.println(tokens.size());//6 for (AnalyzeResponse.AnalyzeToken token : tokens) { System.out.println(token.getTerm() + " "); } }
结果：
6
提前
祝
大家
春节快乐
春节
快乐
2.调用ik_smart分词
@Test public void analyzer_ik_smart() throws Exception { java.lang.String text = "提前祝大家春节快乐！"; TransportClient client = EsClient.get(); AnalyzeRequest request = (new AnalyzeRequest()).analyzer("ik_smart").text(text); List<AnalyzeResponse.AnalyzeToken> tokens = client.admin().indices().analyze(request).actionGet().getTokens(); System.out.println(tokens.size()); for (AnalyzeResponse.AnalyzeToken token : tokens) { System.out.println(token.getTerm() + " "); } }
结果：
4
提前
祝
大家
春节快乐