Elasticsearch Java REST Client Term Vectors API

上一篇：Elasticsearch Java REST Client 删除、修改
下一篇：Elasticsearch Java REST Client 批量操作（Bulk API）

Term Vectors API

官网地址：戳

词条向量
大概就是统计文档中一些参数值的数量

同步执行

# 查询方式一
@RequestMapping("test")
public TermVectorsResponse test(String id) throws IOException {
     TermVectorsRequest request = new TermVectorsRequest("edu-app-user", "_doc", id);
     request.setFields("name");
     return restHighLevelClient.termvectors(request, RequestOptions.DEFAULT);
 }

# 方式二
@RequestMapping("test1")
public TermVectorsResponse test1() throws IOException {
    XContentBuilder docBuilder = XContentFactory.jsonBuilder();
    docBuilder.startObject().field("name", "川").endObject();
    TermVectorsRequest request = new TermVectorsRequest("edu-app-user", "_doc", docBuilder);
    return restHighLevelClient.termvectors(request, RequestOptions.DEFAULT);
}

异步执行

@RequestMapping("test2")
public void test2() throws IOException {
     XContentBuilder docBuilder = XContentFactory.jsonBuilder();
     docBuilder.startObject().field("id", "28").endObject();
     TermVectorsRequest request = new TermVectorsRequest("edu-app-user", "_doc", docBuilder);
     restHighLevelClient.termvectorsAsync(request, RequestOptions.DEFAULT,new TermVectorsESListen());
 }

# 监听类
package com.wdz.es.config.es;

import com.alibaba.fastjson.JSONObject;
import org.elasticsearch.action.ActionListener;
import org.elasticsearch.client.core.TermVectorsResponse;

public class TermVectorsESListen implements ActionListener<TermVectorsResponse> {
    @Override
    public void onResponse(TermVectorsResponse response) {
        System.out.println("成功结果："+ JSONObject.toJSONString(response));
    }

    @Override
    public void onFailure(Exception e) {
        System.out.println("失败结果："+e.getMessage());
    }
}

其他可选方法

# 设置fieldStatistics为false（默认为true）以省略文档计数、文档频率总和、总词频总和。
request.setFieldStatistics(false); 
# 设置termStatistics为true（默认为false）以显示总词频和文档频率。
request.setTermStatistics(true); 
# 设置positions为false（默认为true）以省略位置的输出
request.setPositions(false); 
# 设置offsets为false（默认为true）以省略偏移量的输出
request.setOffsets(false); 
# 设置payloads为false（默认为true）以省略有效负载的输出
request.setPayloads(false); 

Map<String, Integer> filterSettings = new HashMap<>();
filterSettings.put("max_num_terms", 3);
filterSettings.put("min_term_freq", 1);
filterSettings.put("max_term_freq", 10);
filterSettings.put("min_doc_freq", 1);
filterSettings.put("max_doc_freq", 100);
filterSettings.put("min_word_length", 1);
filterSettings.put("max_word_length", 10);
# 设置filterSettings为根据 tf-idf 分数过滤可以返回的术语
request.setFilterSettings(filterSettings);  
Map<String, String> perFieldAnalyzer = new HashMap<>();
perFieldAnalyzer.put("user", "keyword");
# 设置perFieldAnalyzer以指定与该字段具有的分析器不同的分析器
request.setPerFieldAnalyzer(perFieldAnalyzer);  
# 设置realtime为false（默认为true）以接近实时检索词向量
request.setRealtime(false);
# 设置路由参数 
request.setRouting("routing"); 

结果解析

for (TermVectorsResponse.TermVector tv : response.getTermVectorsList()) {
    String fieldname = tv.getFieldName(); 
    int docCount = tv.getFieldStatistics().getDocCount(); 
    long sumTotalTermFreq =
            tv.getFieldStatistics().getSumTotalTermFreq(); 
    long sumDocFreq = tv.getFieldStatistics().getSumDocFreq(); 
    if (tv.getTerms() != null) {
        List<TermVectorsResponse.TermVector.Term> terms =
                tv.getTerms(); 
        for (TermVectorsResponse.TermVector.Term term : terms) {
            String termStr = term.getTerm(); 
            int termFreq = term.getTermFreq(); 
            int docFreq = term.getDocFreq(); 
            long totalTermFreq = term.getTotalTermFreq(); 
            float score = term.getScore(); 
            if (term.getTokens() != null) {
                List<TermVectorsResponse.TermVector.Token> tokens =
                        term.getTokens(); 
                for (TermVectorsResponse.TermVector.Token token : tokens) {
                    int position = token.getPosition(); 
                    int startOffset = token.getStartOffset(); 
                    int endOffset = token.getEndOffset(); 
                    String payload = token.getPayload(); 
                }
            }
        }
    }
}

{	# 查询的索引
    "index":"edu-app-user",
    # 文档类型名称
    "type":"_doc",
    # 索引的id
    "id":"28",
    "docVersion":7,
    # 无数据的时候为false
    "found":true,
    "tookInMillis":10,
    "termVectorsList":[
        {	
        	# 当前查询的指定参数：name
            "fieldName":"name",
            "fieldStatistics":{
                # 当前字段的字段统计信息 - 文档频率总和
                "sumDocFreq":9,
                # 当前字段的字段统计信息 - 文档计数
                "docCount":2,
                当前字段的字段统计信息 - 总词频的总和
                "sumTotalTermFreq":10
            },
            # 当前字段的术语
            "terms":[
                {	# 术语的名称
                    "term":"川",
                    # 术语的术语频率
                    "termFreq":1,
                    # 术语的文档频率
                    "docFreq":null,
                    # 术语的总术语频率
                    "totalTermFreq":null,
                    "score":null,
                    # 术语的标记
                    "tokens":[
                        {   # 令牌的起始偏移量
                            "startOffset":3,
                            # 令牌的结束偏移量
                            "endOffset":4,
                             # 令牌的位置
                            "position":3,
                            #  令牌的有效载荷
                            "payload":null
                        }
                    ]
                },
                {
                    "term":"州",
                    "termFreq":1,
                    "docFreq":null,
                    "totalTermFreq":null,
                    "score":null,
                    "tokens":[
                        {
                            "startOffset":1,
                            "endOffset":2,
                            "position":1,
                            "payload":null
                        }
                    ]
                }
            ]
        }
    ]
}

上一篇：Elasticsearch Java REST Client 删除、修改
下一篇：Elasticsearch Java REST Client 批量操作（Bulk API）