ElasticSearch纠错和自动补全

ElasticSearch纠错和自动补全

1. 建立映射

   PUT /suggest
   {
     "mappings": {
       "properties": {
         "suggest": {
           "type": "completion"
         }
       }
     }
   }
   

   先建立一个mapping,注意类型"type": "completion"

2. 添加一些数据

   POST /suggest/_doc
   {
     "suggest":"数据"
   }
   POST /suggest/_doc
   {
     "suggest":"数据挖掘"
   }
   POST /suggest/_doc
   {
     "suggest":"数据挖掘分析"
   }
   POST /suggest/_doc
   {
     "suggest":"数据挖掘分析工具"
   }
   

   为了方便说明添加一些简单的关键词

3. 示例

   #自动完成
   POST /suggest/_search
   {
     "suggest": {
       "YOUR_SUGGESTION": {
         "prefix": "数据",
         "completion": {
           "field": "suggest",
           "size": 10
         }
       }
     }
   }
   

   自动完成：

   prefix前缀,默认匹配至少第一个字

   field字段名

   size返回数量

   参考：https://www.elastic.co/guide/en/elasticsearch/reference/7.10/search-suggesters.html#completion-suggester

   #术语纠错
   POST /suggest/_search
   {
     "suggest": {
       "YOUR_SUGGESTION": {
         "text": "数据哇掘",
         "term": {
           "suggest_mode": "always",
           "min_word_length": 2,
           "size": 10,
           "field": "suggest"
         }
       }
     }
   }
   

   注意术语纠错的mapping字段不能是text，因为是中文所以进行不能分词不然"数据哇掘"会被分成"数据"、“哇”、"掘"会进行分别纠错导致的结果就没有出错，而不会对"数据哇掘"进行整体纠错。

   min_word_length表示文本长度最少是2，因为中文一个字没法纠错

   参考：https://www.elastic.co/guide/en/elasticsearch/reference/7.10/search-suggesters.html#term-suggester