ElasticSearch使用completion实现补全功能

最新推荐文章于 2024-08-14 06:49:44 发布

王哲晓

最新推荐文章于 2024-08-14 06:49:44 发布

阅读量1.1w

点赞数 7

分类专栏： ElasticSearch

ElasticSearch 专栏收录该内容

5 篇文章 0 订阅

订阅专栏

文章目录

最终效果

实现的最终效果如下图百度搜索相似，输入词的时候返回提示。同时输入拼音也会有相同的提示效果。

completion使用

安装ik中文分词器：https://github.com/medcl/elasticsearch-analysis-ik

安装拼音分词器：https://github.com/medcl/elasticsearch-analysis-pinyin

定义关键词索引并自定义ik+pinyin分词器，要完成补全搜索，必须要用到特殊的数据类型completion，要汉字拼音都能补全，必须要使用自定义的ik+pinyin分词器。

PUT suggest
{
  "settings": {
    "number_of_replicas": 0,
    "number_of_shards": 1,
    "analysis": {
        "analyzer": {
            "ik_pinyin_analyzer": {
                "type": "custom",
                "tokenizer": "ik_max_word",
                "filter": ["my_pinyin", "word_delimiter"]
            }
        },
        "filter": {
            "my_pinyin": {
                "type": "pinyin",
                "first letter": "prefix",
                "padding_char": " "
            }
        }
    }    
  },
  "mappings": {
      "suggest": {
          "properties": {
              "keyword": {
                  "type": "completion",
                  "analyzer": "ik_pinyin_analyzer",
                  "fields": {
                      "key": {
                          "type": "keyword"
                      }
                  }
              },
              "id": {
                  "type": "keyword"
              },
              "createDate": {
                  "type": "date",
                  "format": "yyyy-MM-dd HH:mm:ss"
              }
          }
      }
  }    
}

初始化部分数据

POST _bulk/?refresh=true
{ "index": { "_index": "suggest", "_type": "suggest" }}
{ "keyword": "项目"}
{ "index": { "_index": "suggest", "_type": "suggest" }}
{ "keyword": "项目进度"}
{ "index": { "_index": "suggest", "_type": "suggest" }}
{ "keyword": "项目管理"}
{ "index": { "_index": "suggest", "_type": "suggest" }}
{ "keyword": "项目进度及调整 汇总.doc_文档"}
{ "index": { "_index": "suggest", "_type": "suggest" }}
{ "keyword": "项目"}

使用suggest获取搜索补全建议，并对同一词语去重。

GET /suggest/_search
{
    "suggest": {
        "my-suggest": {
            "prefix": "项目",
            "completion": {
                "field": "keyword",
                "size": 20,
                "skip_duplicates": true
            }
        }
    }
}

Java实现搜索补全代码

/**
 * 获取相关搜索，最多返回9条
 * @param key
 * @return
 */
 public JSONObject getSearchSuggest(String key) {
     CompletionSuggestionBuilder suggestion = SuggestBuilders
             .completionSuggestion("keyword").prefix(key).size(20).skipDuplicates(true);
     SuggestBuilder suggestBuilder = new SuggestBuilder();
     suggestBuilder.addSuggestion("suggest", suggestion);
     SearchResponse response = template.suggest(suggestBuilder, EsConstants.SUGGEST);
     Suggest suggest = response.getSuggest();

     Set<String> keywords = null;
     if (suggest != null) {
         keywords = new HashSet<>();
         List<? extends Suggest.Suggestion.Entry<? extends Suggest.Suggestion.Entry.Option>> entries = suggest.getSuggestion("suggest").getEntries();

         for (Suggest.Suggestion.Entry<? extends Suggest.Suggestion.Entry.Option> entry: entries) {
             for (Suggest.Suggestion.Entry.Option option: entry.getOptions()) {
                 /** 最多返回9个推荐，每个长度最大为20 */
                 String keyword = option.getText().string();
                 if (!StringUtils.isEmpty(keyword) && keyword.length() <= 20) {
                     /** 去除输入字段 */
                     if (keyword.equals(key)) continue;
                     keywords.add(keyword);
                     if (keywords.size() >= 9) {
                         break;
                     }
                 }
             }
         }
     }
     return ApiResult.OK(keywords, "获取推荐词组成功");
 }

热门搜索推荐

上面自定义的索引中，通过fields属性专门存储了类型为keyword的字段keyword.key。可以通过统计keyword.key来获取搜索次数最多的句子。下面是java实现方式

public JSONObject searchHot(Map<String, Object> map) {
    Integer size = 10;
    if (!StringUtils.isEmpty(map.get("size"))) {
        size = (Integer)map.get("size");
    }

    /** 获取最近一个月时间 */
    String preMonth = LocalDateTime.now().minusMonths(1).format(EsConstants.fomatter);
    String now = LocalDateTime.now().format(EsConstants.fomatter);

    /** 统计最近一个月的热门搜索，长度最大10，方便显示 */
    SearchRequestBuilder requestBuilder = template.getClient().prepareSearch(EsConstants.SUGGEST)
            .setQuery(QueryBuilders.rangeQuery("createDate").get(preMonth).lte(now));

    SearchResponse searchResponse = requestBuilder.addAggregation(AggregationBuilders
            .terms("hotSearch").field("keyword.key").size(size)).execute().actionGet();

    Aggregations aggregations = searchResponse.getAggregations();
    Set<String> keywords = null;
    if (aggregations != null) {
        keywords = new HashSet<>();
        Terms hotSearch = aggregations.get("hotSearch");
        List<? extends Terms.Bucket> buckets = hotSearch.getBuckets();
        for (Terms.Bucket bucket: buckets) {
            if (bucket.getKey().toString().length() <= 10) {
                 keywords.add((String)bucket.getKey());
            }
        }
    }
    return ApiResult.OK(keywords, "热门搜索获取成功");
}

原文链接：http://www.pianshen.com/article/7284152357/