› › ›

采样器聚合

一种过滤聚合，用于将任何子聚合的处理限制在得分最高的文档样本中。

示例用例

将分析的重点集中在高相关性匹配上，而不是潜在的非常长的低质量匹配尾部。
降低仅使用样本即可产生有用结果的聚合的运行成本，例如 significant_terms。

示例

在 StackOverflow 数据上查询常用术语 javascript 或较少见的术语 kibana 将匹配许多文档 - 其中大多数缺少 Kibana 一词。为了将 significant_terms 聚合集中在更有可能匹配我们查询中最有趣部分的高分文档上，我们使用一个样本。

resp = client.search(
    index="stackoverflow",
    size="0",
    query={
        "query_string": {
            "query": "tags:kibana OR tags:javascript"
        }
    },
    aggs={
        "sample": {
            "sampler": {
                "shard_size": 200
            },
            "aggs": {
                "keywords": {
                    "significant_terms": {
                        "field": "tags",
                        "exclude": [
                            "kibana",
                            "javascript"
                        ]
                    }
                }
            }
        }
    },
)
print(resp)

response = client.search(
  index: 'stackoverflow',
  size: 0,
  body: {
    query: {
      query_string: {
        query: 'tags:kibana OR tags:javascript'
      }
    },
    aggregations: {
      sample: {
        sampler: {
          shard_size: 200
        },
        aggregations: {
          keywords: {
            significant_terms: {
              field: 'tags',
              exclude: [
                'kibana',
                'javascript'
              ]
            }
          }
        }
      }
    }
  }
)
puts response

const response = await client.search({
  index: "stackoverflow",
  size: 0,
  query: {
    query_string: {
      query: "tags:kibana OR tags:javascript",
    },
  },
  aggs: {
    sample: {
      sampler: {
        shard_size: 200,
      },
      aggs: {
        keywords: {
          significant_terms: {
            field: "tags",
            exclude: ["kibana", "javascript"],
          },
        },
      },
    },
  },
});
console.log(response);

POST /stackoverflow/_search?size=0
{
  "query": {
    "query_string": {
      "query": "tags:kibana OR tags:javascript"
    }
  },
  "aggs": {
    "sample": {
      "sampler": {
        "shard_size": 200
      },
      "aggs": {
        "keywords": {
          "significant_terms": {
            "field": "tags",
            "exclude": [ "kibana", "javascript" ]
          }
        }
      }
    }
  }
}

响应

{
  ...
  "aggregations": {
    "sample": {
      "doc_count": 200, 
      "keywords": {
        "doc_count": 200,
        "bg_count": 650,
        "buckets": [
          {
            "key": "elasticsearch",
            "doc_count": 150,
            "score": 1.078125,
            "bg_count": 200
          },
          {
            "key": "logstash",
            "doc_count": 50,
            "score": 0.5625,
            "bg_count": 50
          }
        ]
      }
    }
  }
}

总共采样了 200 个文档。因此，执行嵌套 significant_terms 聚合的成本是有限的，而不是无限制的。

如果没有 sampler 聚合，请求查询会考虑低质量匹配的完整“长尾”，因此会识别出不太重要的术语，例如 jquery 和 angular，而不是关注更具洞察力的 Kibana 相关术语。

resp = client.search(
    index="stackoverflow",
    size="0",
    query={
        "query_string": {
            "query": "tags:kibana OR tags:javascript"
        }
    },
    aggs={
        "low_quality_keywords": {
            "significant_terms": {
                "field": "tags",
                "size": 3,
                "exclude": [
                    "kibana",
                    "javascript"
                ]
            }
        }
    },
)
print(resp)

response = client.search(
  index: 'stackoverflow',
  size: 0,
  body: {
    query: {
      query_string: {
        query: 'tags:kibana OR tags:javascript'
      }
    },
    aggregations: {
      low_quality_keywords: {
        significant_terms: {
          field: 'tags',
          size: 3,
          exclude: [
            'kibana',
            'javascript'
          ]
        }
      }
    }
  }
)
puts response

const response = await client.search({
  index: "stackoverflow",
  size: 0,
  query: {
    query_string: {
      query: "tags:kibana OR tags:javascript",
    },
  },
  aggs: {
    low_quality_keywords: {
      significant_terms: {
        field: "tags",
        size: 3,
        exclude: ["kibana", "javascript"],
      },
    },
  },
});
console.log(response);

POST /stackoverflow/_search?size=0
{
  "query": {
    "query_string": {
      "query": "tags:kibana OR tags:javascript"
    }
  },
  "aggs": {
    "low_quality_keywords": {
      "significant_terms": {
        "field": "tags",
        "size": 3,
        "exclude": [ "kibana", "javascript" ]
      }
    }
  }
}

响应

{
  ...
  "aggregations": {
    "low_quality_keywords": {
      "doc_count": 600,
      "bg_count": 650,
      "buckets": [
        {
          "key": "angular",
          "doc_count": 200,
          "score": 0.02777,
          "bg_count": 200
        },
        {
          "key": "jquery",
          "doc_count": 200,
          "score": 0.02777,
          "bg_count": 200
        },
        {
          "key": "logstash",
          "doc_count": 50,
          "score": 0.0069,
          "bg_count": 50
        }
      ]
    }
  }
}