# match_phrase 的 slop
# 语法
GET /forum/article/_search
{
"query": {
"match_phrase": {
"content": {
"query": "java spark",
"slop": 1
}
}
}
}
1
2
3
4
5
6
7
8
9
10
11
2
3
4
5
6
7
8
9
10
11
# slop 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
举个例子：
这样一段文本：hello world, java is very good, spark is also very good.
使用 match_phrase 搜索 java spark 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
java --> spark
java --> spark
上面展示了，当固定第一个 term 的时候，后面的 teram 经过移动直到匹配上搜索词的经过
这个移动的次数就是 slop
1
2
3
4
5
6
7
8
9
10
11
12
13
2
3
4
5
6
7
8
9
10
11
12
13
TIP
slop 只指最大移动次数
# 验证 slop
GET /forum/article/_search
{
"query": {
"match_phrase": {
"content": {
"query": "java spark",
"slop": 2
}
}
}
}
1
2
3
4
5
6
7
8
9
10
11
2
3
4
5
6
7
8
9
10
11
响应结果
{
"took": 1,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"failed": 0
},
"hits": {
"total": 2,
"max_score": 1.1324264,
"hits": [
{
"_index": "forum",
"_type": "article",
"_id": "7",
"_score": 1.1324264,
"_source": {
"content": "java spark are very related, because scala is spark's programming language and scala is also based on jvm like java."
}
},
{
"_index": "forum",
"_type": "article",
"_id": "8",
"_score": 0.21395226,
"_source": {
"content": "java are spark very related, because scala is spark's programming language and scala is also based on jvm like java."
}
}
]
}
}
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
尝试着把 slop 的数值调整大一点，之前有好多条数据中都包含了 java 和 spark， 你会发现靠得越近的（slop 相对小的）得分会越高
其实，加了 slop 的 phrase match，就是 proximity match（近似匹配）
- java spark，短语，doc，phrase match
- java spark，可以有一定的距离，但是靠的越近，越先搜索出来，proximity match