elasticsearch玩转大数据大数据 爬虫Python AI Sql

十七、Elasticsearch使用原生cross-fields

2017-07-16  本文已影响33人  编程界的小学生

1、前两篇分别讲了使用使用most-fields和copy_to来解决cross-fields产生的问题。这篇讲解使用ES自带的cross-fields。

2、直接看案例

GET /forum/article/_search
{
  "query": {
    "multi_match": {
      "query": "Peter Smith",
      "type": "cross_fields", 
      "operator" : "and",
      "fields": ["author_first_name", "author_last_name"]
    }
  }
}

注意带上operator:and这句话

结果:

{
  "took": 7,
  "timed_out": false,
  "_shards": {
    "total": 5,
    "successful": 5,
    "failed": 0
  },
  "hits": {
    "total": 2,
    "max_score": 0.5753642,
    "hits": [
      {
        "_index": "forum",
        "_type": "article",
        "_id": "1",
        "_score": 0.5753642,
        "_source": {
          "articleID": "XHDK-A-1293-#fJ3",
          "userID": 1,
          "hidden": false,
          "postDate": "2017-01-01",
          "tag": [
            "java",
            "hadoop"
          ],
          "tag_cnt": 2,
          "view_cnt": 30,
          "title": "this is java and elasticsearch blog",
          "content": "i like to write best elasticsearch article",
          "sub_title": "learning more courses",
          "author_first_name": "Peter",
          "author_last_name": "Smith",
          "new_author_last_name": "Smith",
          "new_author_first_name": "Peter"
        }
      },
      {
        "_index": "forum",
        "_type": "article",
        "_id": "5",
        "_score": 0.51623213,
        "_source": {
          "articleID": "DHJK-B-1395-#Ky5",
          "userID": 3,
          "hidden": false,
          "postDate": "2017-03-01",
          "tag": [
            "elasticsearch"
          ],
          "tag_cnt": 1,
          "view_cnt": 10,
          "title": "this is spark blog",
          "content": "spark is best big data solution based on scala ,an programming language similar to java",
          "sub_title": "haha, hello world",
          "author_first_name": "Tonny",
          "author_last_name": "Peter Smith",
          "new_author_last_name": "Peter Smith",
          "new_author_first_name": "Tonny"
        }
      }
    ]
  }
}

结果发现,要求每个term都必须在任何一个field中出现。
Peter,Smith
要求Peter必须在author_first_name或author_last_name中出现
要求Smith必须在author_first_name或author_last_name

原来most_fiels,可能像Smith Williams也可能会出现,因为most_fields要求只是任何一个field匹配了就可以,匹配的field越多,分数越高
这就解决了前两篇的问题1:尽可能多的field匹配的doc,而不是某个field完全匹配的doc

问题2:most_fields,没办法用minimum_should_match去掉长尾数据,就是匹配的特别少的结果

既然每个term都要求出现,长尾肯定被去除掉了。

问题3:TF/IDF算法,比如Peter Smith和Smith Williams,搜索Peter Smith的时候,由于first_name中很少有Smith的,所以query在所有document中的频率很低,得到的分数很高,可能Smith Williams反而会排在Peter Smith前面

解决:计算IDF的时候,将每个query在每个field中的IDF都取出来,取最小值,就不会出现极端情况下的极大值了

Peter Smith

Peter
Smith

Smith,在author_first_name这个field中,在所有doc的这个Field中,出现的频率很低,导致IDF分数很高;Smith在所有doc的author_last_name field中的频率算出一个IDF分数,因为一般来说last_name中的Smith频率都较高,所以IDF分数是正常的,不会太高;然后对于Smith来说,会取两个IDF分数中,较小的那个分数。就不会出现IDF分过高的情况。

若有兴趣,欢迎来加入群,【Java初学者学习交流群】:458430385,此群有Java开发人员、UI设计人员和前端工程师。有问必答,共同探讨学习,一起进步!
欢迎关注我的微信公众号【Java码农社区】,会定时推送各种干货:


qrcode_for_gh_577b64e73701_258.jpg
上一篇下一篇

猜你喜欢

热点阅读