Elasticsearch搜索推荐

柔光的暖阳◎ 2023-06-22 14:53 9阅读 0赞

> 本文基于Elasticsearch7.x版本.

### 什么是搜索推荐? ###

当我们在Google上搜索信息时, Google会推荐一些相关联的搜索词, 即使我们拼写错了关键词, Google仍然会推荐出我们想要的内容, 这就是搜索推荐.

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2xpdGlhbnhpYW5nX2thb2xh_size_16_color_FFFFFF_t_70]

![在这里插入图片描述][watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2xpdGlhbnhpYW5nX2thb2xh_size_16_color_FFFFFF_t_70 1]

### Elasticsearch Suggesters ###

Elasticsearch Suggesters 包含四个功能, 下面分别介绍这四个功能及其Java Client的使用.

![在这里插入图片描述][20191212170850254.png]

### Term Suggester ###

Term Suggester 即推荐相似的term. 原理是对输入的搜索文本进行分词得到term, 然后再去倒排索引中找到相似的term返回, 如果存在相同的term, 就返回为空.

**语法**

POST 索引名/_search
    {
      "suggest": {
        "任意名称": {
          "text": "搜索文本",
          "term": {
            "field": "字段",
            "analyzer": "",
            "size": "",
            "sort": "",
            "suggest_mode": "",
            "prefix_length":"",
            ...
          }
        }
      }
    }

(1) analyzer

搜索文本使用的分词器.

(2) size

最多返回几个推荐term.

(3) sort

推荐term的排序.

*  score  
    相关度分数优先.
 *  frequency  
    相似term的个数优先.

(4) suggest\_mode

term的推荐规则.

*  missing  
    如果倒排索引中存在相同的term就不推荐, 默认配置.
 *  popular  
    只推荐比原来term出现频率更高的term.
 *  always  
    不管倒排索引中是否存在相同的term都要推荐.(当然如果没有相似的term就返回空)

(5) prefix\_length

匹配前缀长度, 默认为1. 默认情况下必须匹配首字母才会进行推荐.

(6) 其他

此外还有很多参数, 可以去官网上查看.

> https://www.elastic.co/guide/en/elasticsearch/reference/current/search-suggesters.html\#term-suggester

**实例**

(1) 添加数据

POST articles/_bulk
    { "index" : { } }
    { "title": "lucene", "content": "lucene is very cool"}
    { "index" : { } }
    { "title": "elasticsearch", "content": "Elasticsearch builds on top of lucene"}
    { "index" : { } }
    { "title": "Elasticsearch", "content": "Elasticsearch rock"}
    { "index" : { } }
    { "title": "ELK", "content": "elastic is the company behind ELK stack"}
    { "index" : { } }
    { "title": "Elk", "content": "Elk stack rocks"}
    { "index" : {} }
    { "title": "es", "content": "elasticsearch is rock solid"}

(2) 实例一

POST articles/_search
    {
      "suggest": {
        "my_suggestion_01": {
          "text": "lucen rocks",
          "term": {
            "field": "content"
          }
        },
        "my_suggestion_02": {
          "text": "wucene",
          "term": {
            "prefix_length": 0,
            "field": "title"
          }
        }
      }
    }

结果:

"suggest" : {
    	"my_suggestion_01" : [
    	  {
    	    "text" : "lucen",
    	    "offset" : 0,
    	    "length" : 5,
    	    "options" : [
    	      {
    	        "text" : "lucene",
    	        "score" : 0.8,
    	        "freq" : 2
    	      }
    	    ]
    	  },
    	  {
    	    "text" : "rocks",
    	    "offset" : 6,
    	    "length" : 5,
    	    "options" : [ ]
    	  }
    	],
    	"my_suggestion_02" : [
    	  {
    	    "text" : "wucene",
    	    "offset" : 0,
    	    "length" : 6,
    	    "options" : [
    	      {
    	        "text" : "lucene",
    	        "score" : 0.8333333,
    	        "freq" : 1
    	      }
    	    ]
    	  }
    	]
    }

(3) 实例二

POST articles/_search
    {
      "suggest": {
        "text": "lucen rocks",
        "my_suggestion_01": {
          "term": {
    	    "suggest_mode": "always",
               "field": "content"
          }
        },
        "my_suggestion_02": {
          "term": {
            "field": "title"
          }
        }
      }
    }

结果:

"suggest" : {
    	"my_suggestion_01" : [
    	  {
    	    "text" : "lucen",
    	    "offset" : 0,
    	    "length" : 5,
    	    "options" : [
    	      {
    	        "text" : "lucene",
    	        "score" : 0.8,
    	        "freq" : 2
    	      }
    	    ]
    	  },
    	  {
    	    "text" : "rocks",
    	    "offset" : 6,
    	    "length" : 5,
    	    "options" : [
    	      {
    	        "text" : "rock",
    	        "score" : 0.75,
    	        "freq" : 2
    	      }
    	    ]
    	  }
    	],
    	"my_suggestion_02" : [
    	  {
    	    "text" : "lucen",
    	    "offset" : 0,
    	    "length" : 5,
    	    "options" : [
    	      {
    	        "text" : "lucene",
    	        "score" : 0.8,
    	        "freq" : 1
    	      }
    	    ]
    	  },
    	  {
    	    "text" : "rocks",
    	    "offset" : 6,
    	    "length" : 5,
    	    "options" : [ ]
    	  }
    	]
    }

### Phrase Suggester ###

Phrase Suggester 即推荐相似的短语. 它在Term Suggester的基础上, 考量多个term之间的关系(比如是否同时出现在索引的原文里, 相邻程度以及词频等), 推荐一个相似的短语.

**语法**

POST 索引名/_search
    {
      "suggest": {
        "任意名称": {
          "text": "搜索文本",
          "phrase": {
            "field": "字段",
            "analyzer": "",
            "size": "",
            "max_errors": "",
            "highlight":"",
            ...
          }
        }
      }
    }

(1) analyzer

搜索文本使用的分词器.

(2) size

最多返回几个推荐短语, 默认5个, 常用的值为3或5.

(3) max\_errors

最多可以拼写错的term数, 默认1, 常用值为1或2.

(4) highlight

将短语中推荐的term高亮显示.

(5) 其他

此外还有很多参数, 可以去官网上查看.

> https://www.elastic.co/guide/en/elasticsearch/reference/current/search-suggesters.html\#phrase-suggester

**实例**

(1) 添加数据

POST articles/_bulk
    { "index" : { } }
    { "title": "lucene", "content": "lucene is very cool"}
    { "index" : { } }
    { "title": "elasticsearch", "content": "Elasticsearch builds on top of lucene"}
    { "index" : { } }
    { "title": "Elasticsearch", "content": "Elasticsearch rocks"}
    { "index" : { } }
    { "title": "ELK", "content": "elastic is the company behind ELK stack"}
    { "index" : { } }
    { "title": "Elk", "content": "Elk stack rocks"}

(2) 搜索推荐

POST articles/_search
 {
 "suggest": {
 "my_suggestion": {
 "text": "lucne and elasticsear rock hello world",
 "phrase": {
 "field": "content",
 "max_errors": 2,
 "highlight": {
 "pre_tag": "",
 "post_tag": ""
 }
 }
 }
 }
 }

结果:

"suggest" : {
 	"my_suggestion" : [
 	 {
 	 "text" : "lucne and elasticsear rock hello world",
 	 "offset" : 0,
 	 "length" : 38,
 	 "options" : [
 	 {
 	 "text" : "lucene and elasticsearch rock hello world",
 	 "highlighted" : "lucene and elasticsearch rock hello world",
 	 "score" : 1.5109152E-4
 	 },
 	 {
 	 "text" : "lucne and elasticsearch rocks hello world",
 	 "highlighted" : "lucne and elasticsearch rocks hello world",
 	 "score" : 1.4691541E-4
 	 },
 	 {
 	 "text" : "lucene and elasticsear rocks hello world",
 	 "highlighted" : "lucene and elasticsear rocks hello world",
 	 "score" : 1.4548852E-4
 	 },
 	 {
 	 "text" : "lucne and elasticsearch rock hello world",
 	 "highlighted" : "lucne and elasticsearch rock hello world",
 	 "score" : 1.0102803E-4
 	 },
 	 {
 	 "text" : "lucene and elasticsear rock hello world",
 	 "highlighted" : "lucene and elasticsear rock hello world",
 	 "score" : 1.0004682E-4
 	 }
 	 ]
 	 }
 	]
 }

### Completion Suggester ###

Completion Suggester即自动补全, 当我们输入前缀后会立即推荐完整的字段值. Completion Suggester的实现与Term Suggester和Phrase Suggester不同, 它不是基于倒排索引, 而是基于一种叫FST（finite state transducers）的数据结构.

> 对于需要实现自动补全的字段, 我们需要将其类型定义为completion, 然后ES在创建倒排索引的同时会创建FST. FST保存在内存中, 所以满足自动补全对性能的要求.  
> 注意: FST只支持前缀查找.

> FST的详细介绍: https://www.shenyanchao.cn/blog/2018/12/04/lucene-fst/

**语法**

POST 索引名/_search
    {
      "suggest": {
        "任意名称": {
          "prefix": "搜索文本",
          "completion": {
            "field": "字段",
    		"size": ""
    		"fuzzy": {
    		}
            ...
          }
        }
      }
    }

(1) size

最多返回几个推荐结果.

(2) fuzzy

支持模糊搜索, 自动将拼写错误的搜索文本进行纠正, 然后去匹配字段值. 只能用于text或keyword类型的字段.

*  fuzziness  
    代表最多纠正多少个字母, 默认AUTO.
 *  min\_length  
    搜索词的最小长度, 默认为3. 搜索词的长度小于这个值, 则不会触发模糊搜索.
 *  prefix\_length  
    不会进行错误纠正的最小前缀长度, 默认为1. 默认情况下第一个字符不会被纠正.

(3) 其他

此外还有很多参数, 可以去官网上查看.

> https://www.elastic.co/guide/en/elasticsearch/reference/current/search-suggesters.html\#completion-suggester

**实例**

(1) 定义Mapping

PUT articles
    {
      "mappings": {
        "properties": {
          "title_completion": {
            "type": "completion"
          }
        }
      }
    }

(2) 添加数据

POST articles/_bulk
    { "index" : { } }
    { "title_completion": "lucene is very cool"}
    { "index" : { } }
    { "title_completion": "Elasticsearch builds on top of lucene"}
    { "index" : { } }
    { "title_completion": "Elasticsearch rocks"}
    { "index" : { } }
    { "title_completion": "elastic is the company behind ELK stack"}
    { "index" : { } }
    { "title_completion": "Elk stack rocks"}

(3) 实例一

POST articles/_search
    {
      "suggest": {
        "my_suggestion": {
          "prefix": "elk",
          "completion": {
            "field": "title_completion"
          }
        }
      }
    }

结果:

"suggest" : {
    	"my_suggestion" : [
    	  {
    	    "text" : "elk",
    	    "offset" : 0,
    	    "length" : 3,
    	    "options" : [
    	      {
    	        "text" : "Elk stack rocks",
    	        "_index" : "articles",
    	        "_type" : "_doc",
    	        "_id" : "WZs48G4BY9AP5Vz7khqz",
    	        "_score" : 1.0,
    	        "_source" : {
    	          "title_completion" : "Elk stack rocks"
    	        }
    	      }
    	    ]
    	  }
    	]
    }

(4) 实例二

POST articles/_search
    {
      "suggest": {
        "my_suggestion": {
          "prefix": "elz",
          "completion": {
            "field": "title_completion",
            "fuzzy": {
              "fuzziness": 2
            }
          }
        }
      }
    }

结果:

"suggest" : {
    	"my_suggestion" : [
    	  {
    	    "text" : "elz",
    	    "offset" : 0,
    	    "length" : 3,
    	    "options" : [
    	      {
    	        "text" : "Elasticsearch builds on top of lucene",
    	        "_index" : "articles",
    	        "_type" : "_doc",
    	        "_id" : "Vps48G4BY9AP5Vz7khqz",
    	        "_score" : 1.0,
    	        "_source" : {
    	          "title_completion" : "Elasticsearch builds on top of lucene"
    	        }
    	      },
    	      {
    	        "text" : "Elasticsearch rocks",
    	        "_index" : "articles",
    	        "_type" : "_doc",
    	        "_id" : "V5s48G4BY9AP5Vz7khqz",
    	        "_score" : 1.0,
    	        "_source" : {
    	          "title_completion" : "Elasticsearch rocks"
    	        }
    	      },
    	      {
    	        "text" : "Elk stack rocks",
    	        "_index" : "articles",
    	        "_type" : "_doc",
    	        "_id" : "WZs48G4BY9AP5Vz7khqz",
    	        "_score" : 1.0,
    	        "_source" : {
    	          "title_completion" : "Elk stack rocks"
    	        }
    	      },
    	      {
    	        "text" : "elastic is the company behind ELK stack",
    	        "_index" : "articles",
    	        "_type" : "_doc",
    	        "_id" : "WJs48G4BY9AP5Vz7khqz",
    	        "_score" : 1.0,
    	        "_source" : {
    	          "title_completion" : "elastic is the company behind ELK stack"
    	        }
    	      }
    	    ]
    	  }
    	]
    }

### Context Suggester ###

在Completion Suggester的基础上, 为completion类型的字段添加context mappings, 用于and/or的条件过滤, 这就是Context Suggester.

**语法**

Completion Suggester需要把字段类型定义为completion, Context Suggester在其基础上增加context mappings配置.

> https://www.elastic.co/guide/en/elasticsearch/reference/current/search-suggesters.html\#context-suggester

(1) mapping定义

PUT 索引名/_mapping
    {
      "properties": {
        "字段名":{
          "type": "completion",
          "contexts":[
            {
            "type":"category",
            "name":"任意名称"
            },
            {
            "type":"geo",
            "name":"任意名称"
            }
    	  ]
        }
      }
    }

*  category  
    字段的关联分类.
 *  geo  
    字段的关联地理位置.

(2) 搜索推荐

POST 索引名/_search
    {
      "suggest": {
        "任意名称": {
          "prefix": "搜索文本",
          "completion":{
            "field":"字段",
            "contexts":{
              "上下文字段":"字段值"
            }
          }
        }
      }
    }

**实例**

(1) 定义mapping

PUT articles
    {
      "mappings": {
        "properties": {
          "content": {
            "type": "completion",
            "contexts":[{
              "type":"category",
              "name":"article_category"
            }]
          }
        }
      }
    }

(2) 添加数据

POST articles/_doc
    {
      "content": {
        "input": ["timmy's", "starbucks", "dunkin donuts"],
        "contexts": {
          "article_category": ["cafe", "food"]
        }
      }
    }
    
    POST articles/_doc
    {
      "content": {
        "input": ["I love the star war movies", "star wars"],
        "contexts": {
          "article_category": ["movie"]
        }
      }
    }

(3) 搜索推荐

POST articles/_search
    {
      "suggest": {
        "my_suggestion": {
          "prefix": "star",
          "completion": {
            "field": "content",
            "contexts": {
              "article_category": ["cafe"]
            }
          }
        }
      }
    }

结果:

"suggest" : {
    	"my_suggestion" : [
    	  {
    	    "text" : "star",
    	    "offset" : 0,
    	    "length" : 4,
    	    "options" : [
    	      {
    	        "text" : "starbucks",
    	        "_index" : "articles",
    	        "_type" : "_doc",
    	        "_id" : "J_OH-m4BBkLWhs278x-1",
    	        "_score" : 1.0,
    	        "_source" : {
    	          "content" : {
    	            "input" : [
    	              "timmy's",
    	              "starbucks",
    	              "dunkin donuts"
    	            ],
    	            "contexts" : {
    	              "article_category" : [
    	                "cafe",
    	                "food"
    	              ]
    	            }
    	          }
    	        },
    	        "contexts" : {
    	          "article_category" : [
    	            "cafe"
    	          ]
    	        }
    	      }
    	    ]
    	  }
    	]
    }

前缀为star的搜索结果有两个, 但同时满足类型为cafe的搜索结果只有一个.

### Suggester Java Client ###

下面介绍Elasticsearch Suggester Java Client 的使用, 我们来将上文的实例转化为 Java Client.

**main方法**

public static void main(String[] args) throws IOException {
        RestHighLevelClient client = new RestHighLevelClient(
                RestClient.builder(
                        new HttpHost("localhost", 9200, "http")));
    
    //        bulkIndex(client);
    //        termSuggest(client);
    //        phraseSuggest(client);
    
    //        completionBulkIndex(client);
    //        completionSuggest(client);
    
    //        contextIndex(client);
    //        contextSuggest(client);
    
        client.close();
    }

**Term Suggester Client**

(1) 添加数据

private static void bulkIndex(RestHighLevelClient client) throws IOException {
        BulkRequest request = new BulkRequest();
        request.add(new IndexRequest("articles").id("1")
                .source(XContentType.JSON, "title", "lucene", "content", "lucene is very cool"));
        request.add(new IndexRequest("articles").id("2")
                .source(XContentType.JSON, "title", "elasticsearch", "content", "Elasticsearch builds on top of lucene"));
        request.add(new IndexRequest("articles").id("3")
                .source(XContentType.JSON, "title", "Elasticsearch", "content", "Elasticsearch rock"));
        request.add(new IndexRequest("articles").id("4")
                .source(XContentType.JSON, "title", "ELK", "content", "elastic is the company behind ELK stack"));
        request.add(new IndexRequest("articles").id("5")
                .source(XContentType.JSON, "title", "Elk", "content", "Elk stack rocks"));
        request.add(new IndexRequest("articles").id("6")
                .source(XContentType.JSON, "title", "es", "content", "elasticsearch is rock solid"));
    
        client.bulk(request, RequestOptions.DEFAULT);
    }

(2) 搜索推荐

private static void termSuggest(RestHighLevelClient client) throws IOException {
        //创建一个search请求
        SearchRequest request = new SearchRequest("articles");
    
        //创建两个suggestion
        TermSuggestionBuilder termSuggestionBuilder1 = SuggestBuilders.termSuggestion("content").text("lucen rocks");
        termSuggestionBuilder1.suggestMode(TermSuggestionBuilder.SuggestMode.ALWAYS);
        TermSuggestionBuilder termSuggestionBuilder2 = SuggestBuilders.termSuggestion("title").text("wucene");
        termSuggestionBuilder2.prefixLength(0);
    
        //创建SuggestBuilder, 添加两个termSuggestion
        SuggestBuilder suggestBuilder = new SuggestBuilder();
        suggestBuilder.addSuggestion("my_suggestion_01", termSuggestionBuilder1).addSuggestion("my_suggestion_02", termSuggestionBuilder2);
    
        //加入source
        SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
        sourceBuilder.size(0);
        sourceBuilder.suggest(suggestBuilder);
        request.source(sourceBuilder);
    
        //发送请求
        SearchResponse response = client.search(request, RequestOptions.DEFAULT);
    
        //处理结果
        if (RestStatus.OK == response.status()) {
            Suggest suggest = response.getSuggest();
    
            TermSuggestion termSuggestion01 = suggest.getSuggestion("my_suggestion_01");
            for (TermSuggestion.Entry options : termSuggestion01.getEntries()) {
                for (TermSuggestion.Entry.Option option : options) {
                    String text = option.getText().string();
                    System.out.println("my_suggestion_01推荐的term= " + text);
                }
            }
    
            TermSuggestion termSuggestion02 = suggest.getSuggestion("my_suggestion_02");
            for (TermSuggestion.Entry options : termSuggestion02.getEntries()) {
                for (TermSuggestion.Entry.Option option : options) {
                    String text = option.getText().string();
                    System.out.println("my_suggestion_02推荐的term= " + text);
                }
            }
        }
    }

**Phrase Suggester Client**

(1) 添加数据

(2) 搜索推荐

private static void phraseSuggest(RestHighLevelClient client) throws IOException {
 //创建一个search请求
 SearchRequest request = new SearchRequest("articles");
 
 //创建suggestion
 PhraseSuggestionBuilder phraseSuggestionBuilder = SuggestBuilders.phraseSuggestion("content").text("lucne and elasticsear rock hello world").maxErrors(2).highlight("", "");
 
 //创建SuggestBuilder, 添加phraseSuggestion
 SuggestBuilder suggestBuilder = new SuggestBuilder();
 suggestBuilder.addSuggestion("my_suggestion", phraseSuggestionBuilder);
 
 //加入source
 SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
 sourceBuilder.size(0);
 sourceBuilder.suggest(suggestBuilder);
 request.source(sourceBuilder);
 
 //发送请求
 SearchResponse response = client.search(request, RequestOptions.DEFAULT);
 
 //处理结果
 if (RestStatus.OK == response.status()) {
 Suggest suggest = response.getSuggest();
 PhraseSuggestion suggestion = suggest.getSuggestion("my_suggestion");
 for (PhraseSuggestion.Entry options : suggestion.getEntries()) {
 for (PhraseSuggestion.Entry.Option option : options) {
 System.out.println("my_suggestion推荐的phrase= " + option.getText());
 System.out.println("my_suggestion推荐的highlighted phrase= " + option.getHighlighted());
 }
 }
 }
 }

**Completion Suggester Client**

(1) 添加数据

private static void completionBulkIndex(RestHighLevelClient client) throws IOException {
        //定义mapping
        CreateIndexRequest indexRequest = new CreateIndexRequest("articles");
        XContentBuilder builder = jsonBuilder()
                .startObject()
                    .startObject("properties")
                        .startObject("title_completion").field("type", "completion")
                        .endObject()
                    .endObject()
                .endObject();
        indexRequest.mapping(builder);
    
        client.indices().create(indexRequest, RequestOptions.DEFAULT);
    
        //批量添加数据
        BulkRequest bulkRequest = new BulkRequest();
        bulkRequest.add(new IndexRequest("articles").id("1")
                .source(XContentType.JSON, "title_completion", "lucene is very cool"));
        bulkRequest.add(new IndexRequest("articles").id("2")
                .source(XContentType.JSON, "title_completion", "Elasticsearch builds on top of lucene"));
        bulkRequest.add(new IndexRequest("articles").id("3")
                .source(XContentType.JSON, "title_completion", "Elasticsearch rock"));
        bulkRequest.add(new IndexRequest("articles").id("4")
                .source(XContentType.JSON, "title_completion", "elastic is the company behind ELK stack"));
        bulkRequest.add(new IndexRequest("articles").id("5")
                .source(XContentType.JSON, "title_completion", "Elk stack rocks"));
        bulkRequest.add(new IndexRequest("articles").id("6")
                .source(XContentType.JSON, "title_completion", "elasticsearch is rock solid"));
    
        client.bulk(bulkRequest, RequestOptions.DEFAULT);
    }

(2) 搜索推荐

private static void completionSuggest(RestHighLevelClient client) throws IOException {
        //创建一个search请求
        SearchRequest request = new SearchRequest("articles");
    
        //创建suggestion
        CompletionSuggestionBuilder completionSuggestionBuilder = SuggestBuilders.completionSuggestion("title_completion").prefix("elk");
    
        //创建SuggestBuilder, 添加completionSuggestion
        SuggestBuilder suggestBuilder = new SuggestBuilder();
        suggestBuilder.addSuggestion("my_suggestion", completionSuggestionBuilder);
    
        //加入source
        SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
        sourceBuilder.size(0);
        sourceBuilder.suggest(suggestBuilder);
        request.source(sourceBuilder);
    
        //发送请求
        SearchResponse response = client.search(request, RequestOptions.DEFAULT);
    
        //处理结果
        if (RestStatus.OK == response.status()) {
            Suggest suggest = response.getSuggest();
            CompletionSuggestion suggestion = suggest.getSuggestion("my_suggestion");
            for (CompletionSuggestion.Entry options : suggestion.getEntries()) {
                for (CompletionSuggestion.Entry.Option option : options) {
                    Text text = option.getText();
                    System.out.println("my_suggestion推荐的completion suggestion= " + text);
                }
            }
        }
    }

**context Suggester Client**

(1) 添加数据

private static void contextIndex(RestHighLevelClient client) throws IOException {
        //定义mapping
        CreateIndexRequest createIndexRequest = new CreateIndexRequest("articles");
        XContentBuilder builder = jsonBuilder()
                .startObject()
                    .startObject("properties")
                        .startObject("content")
                                .field("type", "completion")
                                .field("contexts").startArray().startObject()
                                    .field("type","category")
                                    .field("name","article_category")
                                    .endObject().endArray()
                        .endObject()
                    .endObject()
                .endObject();
        createIndexRequest.mapping(builder);
    
        client.indices().create(createIndexRequest, RequestOptions.DEFAULT);
    
        //添加数据
        IndexRequest indexRequest1 = new IndexRequest("articles");
        XContentBuilder builder1 = jsonBuilder()
                .startObject()
                    .startObject("content")
                            .field("input")
                                .startArray().value("timmy's").value("starbucks").value("dunkin donuts")
                                .endArray()
                            .field("contexts")
                                .startObject().field("article_category")
                                    .startArray().value("cafe").value("food")
                                    .endArray()
                                .endObject()
                    .endObject()
                .endObject();
        indexRequest1.source(builder1);
        client.index(indexRequest1, RequestOptions.DEFAULT);
    
        IndexRequest indexRequest2 = new IndexRequest("articles");
        XContentBuilder builder2 = jsonBuilder()
                .startObject()
                    .startObject("content")
                        .field("input")
                            .startArray().value("I love the star war movies").value("star wars")
                            .endArray()
                        .field("contexts")
                            .startObject().field("article_category")
                                .startArray().value("movie")
                                .endArray()
                             .endObject()
                        .endObject()
                .endObject();
        indexRequest2.source(builder2);
    
        client.index(indexRequest2, RequestOptions.DEFAULT);
    }

(2) 搜索推荐

private static void contextSuggest(RestHighLevelClient client) throws IOException {
 //创建一个search请求
 SearchRequest request = new SearchRequest("articles");
 
 //创建suggestion
 CompletionSuggestionBuilder completionSuggestionBuilder = SuggestBuilders.completionSuggestion("content").prefix("star").contexts(Collections.<String, List<? extends ToXContent>>singletonMap("article_category", Collections.singletonList(CategoryQueryContext.builder().setCategory("cafe").build())));
 
 
 //创建SuggestBuilder, 添加contextSuggestion
 SuggestBuilder suggestBuilder = new SuggestBuilder();
 suggestBuilder.addSuggestion("my_suggestion", completionSuggestionBuilder);
 
 //加入source
 SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
 sourceBuilder.size(0);
 sourceBuilder.suggest(suggestBuilder);
 request.source(sourceBuilder);
 
 //发送请求
 SearchResponse response = client.search(request, RequestOptions.DEFAULT);
 
 //处理结果
 if (RestStatus.OK == response.status()) {
 Suggest suggest = response.getSuggest();
 CompletionSuggestion suggestion = suggest.getSuggestion("my_suggestion");
 for (CompletionSuggestion.Entry options : suggestion.getEntries()) {
 for (CompletionSuggestion.Entry.Option option : options) {
 Text text = option.getText();
 System.out.println("my_suggestion推荐的context suggestion= " + text);
 }
 }
 }
 }

[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2xpdGlhbnhpYW5nX2thb2xh_size_16_color_FFFFFF_t_70]: https://img-blog.csdnimg.cn/20191204214904245.jpg?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2xpdGlhbnhpYW5nX2thb2xh,size_16,color_FFFFFF,t_70
[watermark_type_ZmFuZ3poZW5naGVpdGk_shadow_10_text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2xpdGlhbnhpYW5nX2thb2xh_size_16_color_FFFFFF_t_70 1]: https://img-blog.csdnimg.cn/20191204214855839.jpg?x-oss-process=image/watermark,type_ZmFuZ3poZW5naGVpdGk,shadow_10,text_aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2xpdGlhbnhpYW5nX2thb2xh,size_16,color_FFFFFF,t_70
[20191212170850254.png]: https://img-blog.csdnimg.cn/20191212170850254.png