ES实现搜素建议

1. 什么是搜索建议？

帮助⽤户在搜索的过程，对输入中进行纠错，提示建议性词语。

例如：

2. ES实现原理

输⼊的⽂本分解为 Token，然后在索引的字典里查找相似的 Term 并返回。

3. 对应API，Term& Phrase Suggester

TermSuggester

3种Suggestion Mode

Missing – 如索引中已经存在，就不提供建议

Popular – 推荐出现频率更加⾼的词

Always – ⽆论是否存在，都提供建议

一个排序

默认按照 score 排序，也可以按照“frequency”

首字母限制

默认⾸字⺟不⼀致，就不会匹配推荐。解决方法，将 prefix_length 设置为 0，

Phrase Suggester

Suggest Mode ：missing, popular, always 

Max Errors：最多可以拼错的 Terms 数

Confidence：限制返回结果数，默认为1，只有分数高过设定值时，才会返回。

4. 例子

# 删除之前的Index
DELETE articles
# 插入2个测试数据
POST articles/_bulk
{ "index" : { } }
{ "body": "elasticsearch is good"}
{ "index" : { } }
{ "body": "elasticsearch is ok"}

2. 调用查询api

POST /articles/_search
{
  "size": 1,
  "query": {
    "match": {
      "body": "elasticseach"
    }
  },
  "suggest": {
    "term-suggestion": {
      "text": "elasticseach",
      "term": {
        "suggest_mode": "missing",
        "field": "body"
      }
    }
  }
}

对应结果

{
  "took" : 13,
  "timed_out" : false,
  "_shards" : {
    "total" : 1,
    "successful" : 1,
    "skipped" : 0,
    "failed" : 0
  },
  "hits" : {
    "total" : {
      "value" : 0,
      "relation" : "eq"
    },
    "max_score" : null,
    "hits" : [ ]
  },
  "suggest" : {
    "term-suggestion" : [
      {
        "text" : "elasticseach",
        "offset" : 0,
        "length" : 12,
        "options" : [
          {
            "text" : "elasticsearch",
            "score" : 0.9166667,
            "freq" : 2
          }
        ]
      }
    ]
  }
}

option部分就是对应的推荐词

5. 引用文档

https://www.elastic.co/guide/en/elasticsearch/reference/7.1/search-suggesters.html

https://www.elastic.co/guide/en/elasticsearch/reference/7.1/search-suggesters-term.html

https://www.elastic.co/guide/en/elasticsearch/reference/7.1/search-suggesters-phrase.html