ElasticSearch ik分词器基础知识analyzer

概述

有两种analyzer，你根据自己的需要自己选吧，但是一般是选用ik_max_word
如果是ik_max_word的话: 会将文本做最细粒度的拆分，比如会将“*国歌”拆分为“*,中华人民,中华,华人,人民*,人民,人,民,*,共和,和,国国,国歌”，会穷尽各种可能的组合；

如果使用ik_smart的话 会做最粗粒度的拆分，比如会将“*国歌”拆分为“*,国歌”。
如果我们搜索"*" 搜到吗？ 这样就搜不到的.

结论

一般我们是使用ik_max_word,因为拆分比较细粒度,这样搜索效果会更好.

ik分词器的使用

指定my_type的analyzer 为 ik_max_word

PUT /my_index
{
  "mappings": {
    "my_type": {
      "properties": {
        "text": {
          "type": "text",
          "analyzer": "ik_max_word"
        }
      }
    }
  }
}

添加数据

POST /my_index/my_type/_bulk
{ "index": { "_id": "1"} }
{ "text": "男子偷上万元发红包求交女友 被抓获时仍然单身" }
{ "index": { "_id": "2"} }
{ "text": "16岁少女为结婚“变”22岁 7年后想离婚被法院拒绝" }
{ "index": { "_id": "3"} }
{ "text": "深圳女孩骑车逆行撞奔驰 遭索赔被吓哭(图)" }
{ "index": { "_id": "4"} }
{ "text": "女人对护肤品比对男票好？网友神怼" }
{ "index": { "_id": "5"} }
{ "text": "为什么国内的街道招牌用的都是红黄配？" }

测试一下搜索效果

查看分词效果

GET /my_index/_analyze
{
"text": "男子偷上万元发红包求交女友 被抓获时仍然单身",
"analyzer": "ik_max_word"
}

结果: 下面分词那么多了

{
  "tokens": [
    {
      "token": "男子",
      "start_offset": 0,
      "end_offset": 2,
      "type": "CN_WORD",
      "position": 0
    },
    {
      "token": "偷上",
      "start_offset": 2,
      "end_offset": 4,
      "type": "CN_WORD",
      "position": 1
    },
    {
      "token": "上万",
      "start_offset": 3,
      "end_offset": 5,
      "type": "CN_WORD",
      "position": 2
    },
    {
      "token": "万元",
      "start_offset": 4,
      "end_offset": 6,
      "type": "CN_WORD",
      "position": 3
    },
    {
      "token": "万",
      "start_offset": 4,
      "end_offset": 5,
      "type": "TYPE_CNUM",
      "position": 4
    },
    {
      "token": "元",
      "start_offset": 5,
      "end_offset": 6,
      "type": "COUNT",
      "position": 5
    },
    {
      "token": "发红包",
      "start_offset": 6,
      "end_offset": 9,
      "type": "CN_WORD",
      "position": 6
    },
    {
      "token": "发红",
      "start_offset": 6,
      "end_offset": 8,
      "type": "CN_WORD",
      "position": 7
    },
    {
      "token": "红包",
      "start_offset": 7,
      "end_offset": 9,
      "type": "CN_WORD",
      "position": 8
    },
    {
      "token": "求",
      "start_offset": 9,
      "end_offset": 10,
      "type": "CN_CHAR",
      "position": 9
    },
    {
      "token": "交",
      "start_offset": 10,
      "end_offset": 11,
      "type": "CN_CHAR",
      "position": 10
    },
    {
      "token": "女友",
      "start_offset": 11,
      "end_offset": 13,
      "type": "CN_WORD",
      "position": 11
    },
    {
      "token": "被",
      "start_offset": 14,
      "end_offset": 15,
      "type": "CN_CHAR",
      "position": 12
    },
    {
      "token": "抓获",
      "start_offset": 15,
      "end_offset": 17,
      "type": "CN_WORD",
      "position": 13
    },
    {
      "token": "时",
      "start_offset": 17,
      "end_offset": 18,
      "type": "CN_CHAR",
      "position": 14
    },
    {
      "token": "仍然",
      "start_offset": 18,
      "end_offset": 20,
      "type": "CN_WORD",
      "position": 15
    },
    {
      "token": "单身",
      "start_offset": 20,
      "end_offset": 22,
      "type": "CN_WORD",
      "position": 16
    }
  ]
}

测试查询数据

_score值越高的排前面.

GET /my_index/my_type/_search
{
  "query": {
    "match": {
      "text": "16岁少女结婚好还是单身好？"
    }
  }
}

结果

{
  "took": 13,
  "timed_out": false,
  "_shards": {
    "total": 5,
    "successful": 5,
    "skipped": 0,
    "failed": 0
  },
  "hits": {
    "total": 3,
    "max_score": 2.6093416,
    "hits": [
      {
        "_index": "my_index",
        "_type": "my_type",
        "_id": "2",
        "_score": 2.6093416,
        "_source": {
          "text": "16岁少女为结婚“变”22岁 7年后想离婚被法院拒绝"
        }
      },
      {
        "_index": "my_index",
        "_type": "my_type",
        "_id": "4",
        "_score": 1.3300087,
        "_source": {
          "text": "女人对护肤品比对男票好？网友神怼"
        }
      },
      {
        "_index": "my_index",
        "_type": "my_type",
        "_id": "1",
        "_score": 0.26301134,
        "_source": {
          "text": "男子偷上万元发红包求交女友 被抓获时仍然单身"
        }
      }
    ]
  }
}