进阶-第18__深度探秘搜索技术_基于slop参数实现近似匹配以及原理剖析和相关实验

尝鲜

GET /forum/article/_search {     "query": {         "match_phrase": {             "title": {                 "query": "java spark",                 "slop":  1             }         }     } }   结果： {   "took": 1,   "timed_out": false,   "_shards": {     "total": 5,     "successful": 5,     "failed": 0   },   "hits": {     "total": 0,     "max_score": null,     "hits": []   } }

 

slop（移动）的含义是什么？

 

query string，搜索文本，中的几个term，要经过几次移动才能与一个document匹配，这个移动的次数，就是slop

slop实际移动举例

实际举例，一个query string经过几次移动之后可以匹配到一个document，然后设置slop

 

hello world, java is very good, spark is also very good.

 

java spark，match phrase，搜不到

 

如果我们指定了slop，那么就允许java spark进行移动，来尝试与doc进行匹配

 

java           is               very           good         spark         is

 

java     spark

java        -->        spark                      移动一位

java            -->                        spark             移动两位

java             -->                          spark   移动三位

 

这里的slop，就是3，因为java spark这个短语，spark移动了3次，就可以跟一个doc匹配上了

 

slop的含义，不仅仅是说一个query string terms移动几次，跟一个doc匹配上。而是说，一个query string terms，最多可以移动几次去尝试跟一个doc匹配上

 

slop，设置的是3，那么就ok

 

GET /forum/article/_search

{

    "query": {

        "match_phrase": {

            "title": {

                "query": "spark data",

                "slop":  3

            }

        }

    }

}

 

就可以把刚才那个doc匹配上，那个doc会作为结果返回

 

但是如果slop设置的是2，那么java spark，spark最多只能移动2次，此时跟doc是匹配不上的，那个doc是不会作为结果返回的

 

做实验，验证slop的含义

实验一

GET /forum/article/_search {   "query": {     "match_phrase": {       "content": {         "query": "spark data",         "slop": 3       }     }   } } 结果： {   "took": 1,   "timed_out": false,   "_shards": {     "total": 5,     "successful": 5,     "failed": 0   },   "hits": {     "total": 0,     "max_score": null,     "hits": []   } }

 

实验二

GET /forum/article/_search {   "query": {     "match_phrase": {       "content": {         "query": "spark data",         "slop": 2       }     }   } } 结果 {   "took": 1,   "timed_out": false,   "_shards": {     "total": 5,     "successful": 5,     "failed": 0   },   "hits": {     "total": 0,     "max_score": null,     "hits": []   } }

 

实验三

GET /forum/article/_search {   "query": {     "match_phrase": {       "content": {         "query": "spark data",         "slop": 3       }     }   } } 结果： {   "took": 1,   "timed_out": false,   "_shards": {     "total": 5,     "successful": 5,     "failed": 0   },   "hits": {     "total": 1,     "max_score": 0.21824157,     "hits": [       {         "_index": "forum",         "_type": "article",         "_id": "5",         "_score": 0.21824157,         "_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 spark",           "sub_title": "haha, hello world",           "author_first_name": "Tonny",           "author_last_name": "Peter Smith"         }       }     ]   } }

 

Spark  is  best  big  data  solution based on scala ,an programming language similar to java spark

 

spark data

         --> data  移动一位

             -->  data  移动两位

spark               -->   data  移动三位

实验四增强

GET /forum/article/_search {   "query": {     "match_phrase": {       "content": {         "query": "data spark",         "slop": 5       }     }   } } 结果： {   "took": 1,   "timed_out": false,   "_shards": {     "total": 5,     "successful": 5,     "failed": 0   },   "hits": {     "total": 1,     "max_score": 0.154366,     "hits": [       {         "_index": "forum",         "_type": "article",         "_id": "5",         "_score": 0.154366,         "_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 spark",           "sub_title": "haha, hello world",           "author_first_name": "Tonny",           "author_last_name": "Peter Smith"         }       }     ]   } }

 

 

spark             is                          best        big                data

 

data          spark

-->               data/spark   移动一位

spark          àdata     移动两位

spark             -->                      data     移动三位

spark                                         -->               data    移动四位

spark                                                              -->               data    移动五位

 

slop搜索下，关键词离的越近，relevance score就会越高，做实验说明。。。

{   "took": 4,   "timed_out": false,   "_shards": {     "total": 5,     "successful": 5,     "failed": 0   },   "hits": {     "total": 3,     "max_score": 1.3728157,     "hits": [       {         "_index": "forum",         "_type": "article",         "_id": "2",         "_score": 1.3728157,         "_source": {           "articleID": "KDKE-B-9947-#kL5",           "userID": 1,           "hidden": false,           "postDate": "2017-01-02",           "tag": [             "java"           ],           "tag_cnt": 1,           "view_cnt": 50,           "title": "this is java blog",           "content": "i think java is the best programming language",           "sub_title": "learned a lot of course",           "author_first_name": "Smith",           "author_last_name": "Williams",           "new_author_last_name": "Williams",           "new_author_first_name": "Smith"         }       },       {         "_index": "forum",         "_type": "article",         "_id": "5",         "_score": 0.5753642,         "_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 spark",           "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"         }       },       {         "_index": "forum",         "_type": "article",         "_id": "1",         "_score": 0.28582606,         "_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"         }       }     ]   } }

实验

GET /forum/article/_search {   "query": {     "match_phrase": {       "content": {         "query": "java best",         "slop": 15       }     }   } } 结果 {   "took": 3,   "timed_out": false,   "_shards": {     "total": 5,     "successful": 5,     "failed": 0   },   "hits": {     "total": 2,     "max_score": 0.65380025,     "hits": [       {         "_index": "forum",         "_type": "article",         "_id": "2",         "_score": 0.65380025,         "_source": {           "articleID": "KDKE-B-9947-#kL5",           "userID": 1,           "hidden": false,           "postDate": "2017-01-02",           "tag": [             "java"           ],           "tag_cnt": 1,           "view_cnt": 50,           "title": "this is java blog",           "content": "i think java is the best programming language",           "sub_title": "learned a lot of course",           "author_first_name": "Smith",           "author_last_name": "Williams",           "new_author_last_name": "Williams",           "new_author_first_name": "Smith"         }       },       {         "_index": "forum",         "_type": "article",         "_id": "5",         "_score": 0.07111243,         "_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 spark",           "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"         }       }     ]   } }

 

 

其实，加了slop的phrase match，就是proximity match，近似匹配

 

1、java spark，短语，doc，phrase match

2、java spark，可以有一定的距离，但是靠的越近，越先搜索出来，proximity match

 

 

移动搜索的短语，以达到文档的内容