Elasticsearch搜索建议 #

一、建议类型 #

text

建议类型
├── Term Suggester
│   └── 词项纠错建议
├── Phrase Suggester
│   └── 短语纠错建议
└── Completion Suggester
    └── 自动补全建议

二、Completion Suggester #

2.1 映射配置 #

bash

PUT /products
{
  "mappings": {
    "properties": {
      "name": {
        "type": "text"
      },
      "suggest": {
        "type": "completion",
        "analyzer": "standard",
        "search_analyzer": "standard",
        "preserve_separators": true,
        "preserve_position_increments": true,
        "max_input_length": 50
      }
    }
  }
}

参数说明：

参数	说明	默认值
analyzer	索引分析器	simple
search_analyzer	搜索分析器	同analyzer
preserve_separators	保留分隔符	true
preserve_position_increments	保留位置增量	true
max_input_length	最大输入长度	50

2.2 索引数据 #

bash

POST /products/_doc
{
  "name": "iPhone 15 Pro Max",
  "suggest": {
    "input": [
      "iPhone",
      "iPhone 15",
      "iPhone 15 Pro",
      "iPhone 15 Pro Max"
    ],
    "weight": 10
  }
}

权重说明：

weight值越大，排名越靠前
用于控制建议顺序

2.3 查询建议 #

bash

POST /products/_search
{
  "suggest": {
    "product_suggest": {
      "prefix": "iPh",
      "completion": {
        "field": "suggest",
        "size": 5,
        "skip_duplicates": true,
        "fuzzy": {
          "fuzziness": 1,
          "transpositions": true,
          "min_length": 3,
          "prefix_length": 2
        }
      }
    }
  }
}

响应：

json

{
  "suggest": {
    "product_suggest": [
      {
        "text": "iPh",
        "offset": 0,
        "length": 3,
        "options": [
          {
            "text": "iPhone",
            "_index": "products",
            "_id": "1",
            "_score": 10.0,
            "_source": {
              "name": "iPhone 15 Pro Max"
            }
          }
        ]
      }
    ]
  }
}

2.4 参数说明 #

参数	说明
prefix	前缀查询
field	建议字段
size	返回数量
skip_duplicates	跳过重复
fuzzy	模糊配置

2.5 上下文建议 #

映射配置：

bash

PUT /products
{
  "mappings": {
    "properties": {
      "suggest": {
        "type": "completion",
        "contexts": [
          {
            "name": "brand",
            "type": "category"
          },
          {
            "name": "location",
            "type": "geo",
            "precision": 4
          }
        ]
      }
    }
  }
}

索引数据：

bash

POST /products/_doc
{
  "suggest": {
    "input": ["iPhone", "iPhone 15"],
    "weight": 10,
    "contexts": {
      "brand": "Apple",
      "location": {
        "lat": 40.7128,
        "lon": -74.0060
      }
    }
  }
}

查询建议：

bash

POST /products/_search
{
  "suggest": {
    "product_suggest": {
      "prefix": "iPh",
      "completion": {
        "field": "suggest",
        "contexts": {
          "brand": "Apple"
        }
      }
    }
  }
}

三、Term Suggester #

3.1 基本用法 #

bash

POST /products/_search
{
  "suggest": {
    "text": "iPhne",
    "my_suggestion": {
      "term": {
        "field": "name"
      }
    }
  }
}

响应：

json

{
  "suggest": {
    "my_suggestion": [
      {
        "text": "iPhne",
        "offset": 0,
        "length": 5,
        "options": [
          {
            "text": "iPhone",
            "score": 0.8,
            "freq": 100
          }
        ]
      }
    ]
  }
}

3.2 参数配置 #

bash

POST /products/_search
{
  "suggest": {
    "text": "iPhne",
    "my_suggestion": {
      "term": {
        "field": "name",
        "analyzer": "standard",
        "size": 5,
        "sort": "score",
        "suggest_mode": "missing",
        "max_edits": 2,
        "prefix_length": 1,
        "min_word_length": 4,
        "string_distance": "internal"
      }
    }
  }
}

参数	说明	默认值
analyzer	分析器	字段分析器
size	建议数量	5
sort	排序方式	score
suggest_mode	建议模式	missing
max_edits	最大编辑距离	2
prefix_length	前缀长度	1
min_word_length	最小词长度	4
string_distance	距离算法	internal

3.3 suggest_mode #

模式	说明
missing	只对不存在词提供建议
popular	只对热门词提供建议
always	总是提供建议

3.4 string_distance #

算法	说明
internal	ES内部算法
damerau_levenshtein	Damerau-Levenshtein
levenshtein	Levenshtein
jaro_winkler	Jaro-Winkler
ngram	N-gram

四、Phrase Suggester #

4.1 基本用法 #

bash

POST /products/_search
{
  "suggest": {
    "text": "iPhne fiften",
    "my_suggestion": {
      "phrase": {
        "field": "name",
        "gram_size": 2,
        "real_word_error_likelihood": 0.95,
        "max_errors": 2,
        "collate": {
          "query": {
            "match": {
              "name": "{{suggestion}}"
            }
          }
        }
      }
    }
  }
}

4.2 参数配置 #

参数	说明	默认值
gram_size	n-gram大小	1
real_word_error_likelihood	真词错误概率	0.95
max_errors	最大错误数	1
collate	验证查询	-
size	建议数量	5

4.3 高亮建议 #

bash

POST /products/_search
{
  "suggest": {
    "text": "iPhne fiften",
    "my_suggestion": {
      "phrase": {
        "field": "name",
        "highlight": {
          "pre_tag": "<em>",
          "post_tag": "</em>"
        }
      }
    }
  }
}

五、组合使用 #

5.1 搜索与建议组合 #

bash

POST /products/_search
{
  "query": {
    "match": {
      "name": "iPhne"
    }
  },
  "suggest": {
    "text": "iPhne",
    "my_suggestion": {
      "term": {
        "field": "name"
      }
    }
  }
}

5.2 多建议器 #

bash

POST /products/_search
{
  "suggest": {
    "text": "iPhne",
    "term_suggestion": {
      "term": {
        "field": "name"
      }
    },
    "completion_suggestion": {
      "prefix": "iPh",
      "completion": {
        "field": "suggest"
      }
    }
  }
}

六、实际应用 #

6.1 搜索框自动补全 #

bash

PUT /search_suggestions
{
  "mappings": {
    "properties": {
      "suggest": {
        "type": "completion"
      },
      "category": {
        "type": "keyword"
      }
    }
  }
}

POST /search_suggestions/_doc
{
  "suggest": {
    "input": [
      "iPhone",
      "iPhone 15",
      "iPhone 15 Pro"
    ],
    "weight": 10
  },
  "category": "products"
}

查询：

bash

POST /search_suggestions/_search
{
  "suggest": {
    "search_suggest": {
      "prefix": "iPh",
      "completion": {
        "field": "suggest",
        "size": 10,
        "skip_duplicates": true
      }
    }
  }
}

6.2 拼写纠错 #

bash

POST /products/_search
{
  "query": {
    "match": {
      "name": "iPhne"
    }
  },
  "suggest": {
    "text": "iPhne",
    "did_you_mean": {
      "phrase": {
        "field": "name",
        "max_errors": 2,
        "collate": {
          "query": {
            "match": {
              "name": "{{suggestion}}"
            }
          }
        }
      }
    }
  }
}

七、最佳实践 #

7.1 性能优化 #

text

性能优化建议
├── Completion Suggester
│   ├── 使用合适的分析器
│   └── 控制input数量
├── Term Suggester
│   ├── 设置合理的max_edits
│   └── 使用prefix_length
└── 通用
    ├── 控制size大小
    └── 使用skip_duplicates

7.2 使用场景 #

text

建议器选择
├── 自动补全
│   └── Completion Suggester
├── 单词纠错
│   └── Term Suggester
├── 短语纠错
│   └── Phrase Suggester
└── 组合使用
    └── 多建议器组合

八、总结 #

本章介绍了Elasticsearch搜索建议：

Completion Suggester用于自动补全
Term Suggester用于单词纠错
Phrase Suggester用于短语纠错
可组合使用多种建议器
上下文建议支持过滤
合理配置优化性能

下一步，我们将学习高级特性。