查询建议介绍
查询建议是什么？
查询建议，为用户提供良好的使用体验。主要包括： 拼写检查； 自动建议查询词（自动补全）
ES中查询建议的API
查询建议也是使用_search端点地址。在DSL中suggest节点来定义需要的建议查询
示例1：定义单个建议查询词
POST twitter/_search
{
"query" : {
"match": {
"message": "tring out Elasticsearch"
}
},
"suggest" : { <!-- 定义建议查询 -->
"my-suggestion" : { <!-- 一个建议查询名 -->
"text" : "tring out Elasticsearch", <!-- 查询文本 -->
"term" : { <!-- 使用词项建议器 -->
"field" : "message" <!-- 指定在哪个字段上获取建议词 -->
}
}
}
}
示例2：定义多个建议查询词
POST _search
{
"suggest": {
"my-suggest-1" : {
"text" : "tring out Elasticsearch",
"term" : {
"field" : "message"
}
},
"my-suggest-2" : {
"text" : "kmichy",
"term" : {
"field" : "user"
}
}
}
}
示例3：多个建议查询可以使用全局的查询文本
POST _search
{
"suggest": {
"text" : "tring out Elasticsearch",
"my-suggest-1" : {
"term" : {
"field" : "message"
}
},
"my-suggest-2" : {
"term" : {
"field" : "user"
}
}
}
}
Suggester 介绍
Term suggester
term 词项建议器，对给入的文本进行分词，为每个词进行模糊查询提供词项建议。对于在索引中存在词默认不提供建议词，不存在的词则根据模糊查询结果进行排序后取一定数量的建议词。
常用的建议选项：
| 参数 | 描述 |
|---|---|
| text | 指定搜索文本 |
| field | 获取建议词的搜索字段 |
| analyzer | 指定分词器 |
| size | 每个词返回的最大建议词数 |
| sort | 如何对建议词进行排序，可用选项： score：先按评分排序、再按照文档频率排、term排序 frequency：先按照文档频排序、再按评分、term顺序排 |
| suggest_mode | 建议模式： missing：仅在搜索的词项在索引中不存在时才提供建议词 popular：仅建议文档频率比搜索词项高的词 always：总是提供匹配的建议词 |
示例1：
POST twitter/_search
{
"query" : {
"match": {
"message": "tring out Elasticsearch"
}
},
"suggest" : { <!-- 定义建议查询 -->
"my-suggestion" : { <!-- 一个建议查询名 -->
"text" : "tring out Elasticsearch", <!-- 查询文本 -->
"term" : { <!-- 使用词项建议器 -->
"field" : "message" <!-- 指定在哪个字段上获取建议词 -->
}
}
}
}
phrase suggester
phrase 短语建议，在term的基础上，会考量多个term之间的关系，比如是否同时出现在索引的原文里，相邻程度，以及词频等
示例1：
POST /ftq/_search
{
"query": {
"match_all": {}
},
"suggest" : {
"myss":{
"text": "java sprin boot",
"phrase": {
"field": "title"
}
}
}
}
结果1：
{
"took": 177,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"skipped": 0,
"failed": 0
},
"hits": {
"total": 2,
"max_score": 1,
"hits": [
{
"_index": "ftq",
"_type": "_doc",
"_id": "2",
"_score": 1,
"_source": {
"title": "java spring boot",
"content": "lucene is writerd by java"
}
},
{
"_index": "ftq",
"_type": "_doc",
"_id": "1",
"_score": 1,
"_source": {
"title": "lucene solr and elasticsearch",
"content": "lucene solr and elasticsearch for search"
}
}
]
},
"suggest": {
"myss": [
{
"text": "java sprin boot",
"offset": 0,
"length": 15,
"options": [
{
"text": "java spring boot",
"score": 0.20745796
}
]
}
]
}
}
Completion suggester 自动补全
针对自动补全场景而设计的建议器。此场景下用户每输入一个字符的时候，就需要即时发送一次查询请求到后端查找匹配项，在用户输入速度较高的情况下对后端响应速度要求比较苛刻。因此实现上它和前面两个Suggester采用了不同的数据结构，索引并非通过倒排来完成，而是将analyze过的数据编码成FST和索引一起存放。对于一个open状态的索引，FST会被ES整个装载到内存里的，进行前缀查找速度极快。但是FST只能用于前缀查找，这也是Completion Suggester的局限所在。
为了使用自动补全，索引中用来提供补全建议的字段需特殊设计，字段类型为 completion。
PUT music
{
"mappings": {
"_doc" : {
"properties" : {
"suggest" : { <!-- 用于自动补全的字段 -->
"type" : "completion"
},
"title" : {
"type": "keyword"
}
}
}
}
}
Input 指定输入词 Weight 指定排序值（可选）
PUT music/_doc/1?refresh
{
"suggest" : {
"input": [ "Nevermind", "Nirvana" ],
"weight" : 34
}
}
指定不同的排序值：
PUT music/_doc/1?refresh
{
"suggest" : [
{
"input": "Nevermind",
"weight" : 10
},
{
"input": "Nirvana",
"weight" : 3
}
]}
放入一条重复数据
PUT music/_doc/2?refresh
{
"suggest" : {
"input": [ "Nevermind", "Nirvana" ],
"weight" : 20
}
}
示例1：查询建议根据前缀查询：
POST music/_search?pretty
{
"suggest": {
"song-suggest" : {
"prefix" : "nir",
"completion" : {
"field" : "suggest"
}
}
}
}
结果1：
{
"took": 25,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"skipped": 0,
"failed": 0
},
"hits": {
"total": 0,
"max_score": 0,
"hits": []
},
"suggest": {
"song-suggest": [
{
"text": "nir",
"offset": 0,
"length": 3,
"options": [
{
"text": "Nirvana",
"_index": "music",
"_type": "_doc",
"_id": "2",
"_score": 20,
"_source": {
"suggest": {
"input": [
"Nevermind",
"Nirvana"
],
"weight": 20
}
}
},
{
"text": "Nirvana",
"_index": "music",
"_type": "_doc",
"_id": "1",
"_score": 1,
"_source": {
"suggest": [
"Nevermind",
"Nirvana"
]
}
}
]
}
]
}
}
示例2：对建议查询结果去重
POST music/_search?pretty
{
"suggest": {
"song-suggest" : {
"prefix" : "nir",
"completion" : {
"field" : "suggest",
"skip_duplicates": true
}
} }}
结果2：
{
"took": 4,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"skipped": 0,
"failed": 0
},
"hits": {
"total": 0,
"max_score": 0,
"hits": []
},
"suggest": {
"song-suggest": [
{
"text": "nir",
"offset": 0,
"length": 3,
"options": [
{
"text": "Nirvana",
"_index": "music",
"_type": "_doc",
"_id": "2",
"_score": 20,
"_source": {
"suggest": {
"input": [
"Nevermind",
"Nirvana"
],
"weight": 20
}
}
}
]
}
]
}
}
示例3：查询建议文档存储短语
PUT music/_doc/3?refresh
{
"suggest" : {
"input": [ "lucene solr", "lucene so cool","lucene elasticsearch" ],
"weight" : 20
}
}
PUT music/_doc/4?refresh
{
"suggest" : {
"input": ["lucene solr cool","lucene elasticsearch" ],
"weight" : 10
}
}
查询3：
POST music/_search?pretty
{
"suggest": {
"song-suggest" : {
"prefix" : "lucene s",
"completion" : {
"field" : "suggest" ,
"skip_duplicates": true
}
}
}
}
结果3：
{
"took": 3,
"timed_out": false,
"_shards": {
"total": 5,
"successful": 5,
"skipped": 0,
"failed": 0
},
"hits": {
"total": 0,
"max_score": 0,
"hits": []
},
"suggest": {
"song-suggest": [
{
"text": "lucene s",
"offset": 0,
"length": 8,
"options": [
{
"text": "lucene so cool",
"_index": "music",
"_type": "_doc",
"_id": "3",
"_score": 20,
"_source": {
"suggest": {
"input": [
"lucene solr",
"lucene so cool",
"lucene elasticsearch"
],
"weight": 20
}
}
},
{
"text": "lucene solr cool",
"_index": "music",
"_type": "_doc",
"_id": "4",
"_score": 10,
"_source": {
"suggest": {
"input": [
"lucene solr cool",
"lucene elasticsearch"
],
"weight": 10
}
}
}
]
}
]
}
}
