ES标签搜索并解决评分排序问题

时间:2024-01-27 20:15:27

一、概述

需求:

  最近在做一个新闻项目,有这样一个需求,如下:

  1. 用户根据视频内容手动创建标签,标签个数不限
  2. 在视频详情页提供根据标签推荐视频功能,即按本视频的标签进行搜索,标签匹配多的排在前面,匹配少的排在后面
 
经过分析、调研,以单字段存储标签,尝试了下面的几种方案,这里一并写出
不可行方案:
  1. 字段为keyword类型,数据以数组存储,未找到可实现此功能的检索方式
  2. 字段为text类型,多个标签以空格隔开或者数组存储,使用match搜索,数据评分不准确
  3. 字段为text类型,多个标签以空格隔开或者数组存储,使用match结合match_phrase搜索,数据评分扔不准确
可行方案:

1. 字段为text类型,指定分词器为whitespace,以空格分隔标签

 

"mediaTag" : {
   "type" : "text",
   "analyzer": "whitespace"
}

 

2. 字段为text类型,指定分词器为pattern,指定标签分隔字符,以逗号分隔

 

PUT /es_medias_test2
{
  "settings": {
    "analysis": {
      "analyzer": {
        "comma": {  //自定义分词器名称
          "type": "pattern",
          "pattern": ","
        }
      }
    }
  },
  "mappings": {
    "esmedias": {
      "properties": {
        "mediaTag": {
          "type": "text",
          "analyzer": "comma"
        }
      }
    }
  }
}

 

二、可行方案测试(以可行方案一为例)

2. 创建索引
PUT /es_medias_test2
{
  "settings": {
    "index": {
      "number_of_shards": "1",
      "number_of_replicas": "0"
    }
  },
  "mappings": {
    "esmedias": {
      "properties": {
        "mediaTag" : {
            "type" : "text",
            "analyzer": "whitespace"
          }
      }
    }
  }
}

 

2. 添加数据
POST /es_medias_test2/_bulk
{"create":{"_index":"es_medias_test2","_type":"esmedias","_id":"o3kyp3YB_f4AQBwwbA7Q"}}
{"mediaTag":"美国 英国"}
{"create":{"_index":"es_medias_test2","_type":"esmedias","_id":"lHk0p3YB_f4AQBwwvxBz"}}
{"mediaTag":"英国 美国"}
{"create":{"_index":"es_medias_test2","_type":"esmedias","_id":"-Xk1p3YB_f4AQBwwNRBt"}}
{"mediaTag":"美国 法国 英国"}
{"create":{"_index":"es_medias_test2","_type":"esmedias","_id":"AXlYp3YB_f4AQBww9zDT"}}
{"mediaTag":"china 美国 英国"}
{"create":{"_index":"es_medias_test2","_type":"esmedias","_id":"13k1p3YB_f4AQBwwBxDw"}}
{"mediaTag":"美国 英国 士大夫"}
{"create":{"_index":"es_medias_test2","_type":"esmedias","_id":"PXk1p3YB_f4AQBwwfxGI"}}
{"mediaTag":"美国"}
{"create":{"_index":"es_medias_test2","_type":"esmedias","_id":"G3k1p3YB_f4AQBwwahEM"}}
{"mediaTag":"英国 船"}
{"create":{"_index":"es_medias_test2","_type":"esmedias","_id":"G3lap3YB_f4AQBwwNTEX"}}
{"mediaTag":"china 美国"}
{"create":{"_index":"es_medias_test2","_type":"esmedias","_id":"FXlLp3YB_f4AQBwwUCRf"}}
{"mediaTag":"china 美国 法国"}

 

 
4. 测试
GET /es_medias_test2/_search
{
  "query": {
    "match": {
      "mediaTag": "美国 英国"
    }
  }
}

返回结果符合预期

{
  "took" : 1,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : 10,
    "max_score" : 1.8475795,
    "hits" : [
      {
        "_index" : "es_medias_test2",
        "_type" : "esmedias",
        "_id" : "-Xk1p3YB_f4AQBwwNRBt",
        "_score" : 1.8475795,
        "_source" : {
          "mediaTag" : "美国 法国 英国"
        }
      },
      {
        "_index" : "es_medias_test2",
        "_type" : "esmedias",
        "_id" : "FXlLp3YB_f4AQBwwUCRf",
        "_score" : 1.5141833,
        "_source" : {
          "mediaTag" : "china 美国 法国"
        }
      },
      {
        "_index" : "es_medias_test2",
        "_type" : "esmedias",
        "_id" : "o3kyp3YB_f4AQBwwbA7Q",
        "_score" : 0.66557413,
        "_source" : {
          "mediaTag" : "美国 英国"
        }
      },
      {
        "_index" : "es_medias_test2",
        "_type" : "esmedias",
        "_id" : "xXkyp3YB_f4AQBwwpw6Y",
        "_score" : 0.66557413,
        "_source" : {
          "mediaTag" : "美国 英国"
        }
      },
      {
        "_index" : "es_medias_test2",
        "_type" : "esmedias",
        "_id" : "lHk0p3YB_f4AQBwwvxBz",
        "_score" : 0.66557413,
        "_source" : {
          "mediaTag" : "英国 美国"
        }
      },
      {
        "_index" : "es_medias_test2",
        "_type" : "esmedias",
        "_id" : "13k1p3YB_f4AQBwwBxDw",
        "_score" : 0.5578373,
        "_source" : {
          "mediaTag" : "美国 英国 士大夫"
        }
      },
      {
        "_index" : "es_medias_test2",
        "_type" : "esmedias",
        "_id" : "AXlYp3YB_f4AQBww9zDT",
        "_score" : 0.39778596,
        "_source" : {
          "mediaTag" : "china,美国 英国"
        }
      },
      {
        "_index" : "es_medias_test2",
        "_type" : "esmedias",
        "_id" : "G3k1p3YB_f4AQBwwahEM",
        "_score" : 0.39778596,
        "_source" : {
          "mediaTag" : "英国 船"
        }
      },
      {
        "_index" : "es_medias_test2",
        "_type" : "esmedias",
        "_id" : "PXk1p3YB_f4AQBwwfxGI",
        "_score" : 0.33188638,
        "_source" : {
          "mediaTag" : "美国"
        }
      },
      {
        "_index" : "es_medias_test2",
        "_type" : "esmedias",
        "_id" : "G3lap3YB_f4AQBwwNTEX",
        "_score" : 0.26778817,
        "_source" : {
          "mediaTag" : "china 美国"
        }
      }
    ]
  }
}