ES 查询以匹配查询中尽可能多的单词

Question

我的索引中有几百万个文档。我有一个句子，想检索匹配尽可能多的单词的文档。我只需要搜索一个字段 content

curl -X GET "xxx.com:9200/test/_search?pretty" -H 'Content-Type: application/json' -d'
{
    "query" : {
        "bool" : { "must" : [{"term": {"content": {"value": "popular artworks of Banksy"}}}]
    }}
}
'

我想要文档中的查询词越多越好。如果有一个文件的文本中出现了很多艺术作品、Banksy 和一些流行的 - 它应该被评为高分。此外，是否可以降低比其他词出现频率更高的词的匹配权重？喜欢比 Banksy 更受欢迎。我知道我可以使用 boost。但我不想手动设置这些值。如果可能的话，我希望它有一个隐含的理解。

Answer 1

添加包含索引数据、搜索查询和搜索结果的工作示例。

请参阅 match_phrase query and bool queries 上的 ES 文档以获得详细说明。

索引数据：

{
    "content":"popular popular popular artworks artworks Banksy"
}
{
    "content":"popular artworks Banksy"
}
{
    "content":"popular popular artworks Banksy"
}
{
    "content": "popular artworks Banksy Banksy"
}
{
    "content": "popular popular popular artworks artworks artworks Banksy"
}

搜索查询：

    {
  "query": {
    "bool": {
      "should": [
        {
          "match": {
            "content": "popular artworks of Banksy"
          }
        },
        {
          "match_phrase":{
            "content":"popular artworks Banksy Banksy"
          }
        }
      ]
    }
  }
}

搜索结果：

"hits": [
      {
        "_index": "test1",
        "_type": "_doc",
        "_id": "4",
        "_score": 0.4776722,
        "_source": {
          "content": "popular artworks Banksy Banksy"
        }
      },
      {
        "_index": "test1",
        "_type": "_doc",
        "_id": "5",
        "_score": 0.22413516,
        "_source": {
          "content": "popular popular popular artworks artworks artworks Banksy"
        }
      },
      {
        "_index": "test1",
        "_type": "_doc",
        "_id": "1",
        "_score": 0.22279418,
        "_source": {
          "content": "popular popular popular artworks artworks Banksy"
        }
      },
      {
        "_index": "test1",
        "_type": "_doc",
        "_id": "3",
        "_score": 0.21652403,
        "_source": {
          "content": "popular popular artworks Banksy"
        }
      },
      {
        "_index": "test1",
        "_type": "_doc",
        "_id": "2",
        "_score": 0.21318543,
        "_source": {
          "content": "popular artworks Banksy"
        }
      }
    ]

ES 查询以匹配查询中尽可能多的单词

ES query to match as many words from the query

elasticsearch

elasticsearch-query