【发布时间】：2018-10-12 22:39:09
【问题描述】：
我的索引中有这样的话：“Kem, Kemi, Kemah, Kemer, Kemerburgaz, Kemang, Kembs, Kemnay, Kempley, Kempsey, Kemerovo”。
当我搜索“Kem”时，我希望“Kemi”排在最前面，因为它是最接近的词。 （凯姆 + 我 = 凯米）。但它并没有按照我想要的方式进行。
索引：
{
"settings": {
"number_of_shards": 1,
"analysis": {
"filter": {
"autocomplete_filter": {
"type": "edge_ngram",
"min_gram": 2,
"max_gram": 15
}
},
"analyzer": {
"autocomplete": {
"type": "custom",
"tokenizer": "standard",
"filter": [
"lowercase",
"autocomplete_filter"
]
}
}
}
},
"mappings": {
"_doc": {
"properties": {
"name": {
"fields": {
"keyword": {
"type": "keyword"
}
},
"type": "text",
"similarity": "classic",
"analyzer": "autocomplete",
"search_analyzer": "standard"
},
"id": {
"type": "keyword"
},
"slug": {
"type": "keyword"
},
"type": {
"type": "keyword"
}
}
}
}
}
查询：
{
"from" : 0, "size" : 10,
"query": {
"bool": {
"must": [
{
"match": {
"name": "Kem"
}
}
],
"should": [
{
"term": {
"name.keyword": {
"value": "Kem"
}
}
}
]
}
}
}
'
结果：
{
"took" : 6,
"timed_out" : false,
"_shards" : {
"total" : 1,
"successful" : 1,
"skipped" : 0,
"failed" : 0
},
"hits" : {
"total" : 143,
"max_score" : 20.795834,
"hits" : [
{
"_index" : "destinations",
"_type" : "_doc",
"_id" : "lPL8Y2YBqxTX_xwrZlGc",
"_score" : 20.795834,
"_source" : {
"id" : "c6317201",
"name" : "Kem",
"slug" : "yurtdisi/karelya-cumhuriyeti/kem"
}
},
{
"_index" : "destinations",
"_type" : "_doc",
"_id" : "se78Y2YBqxTX_xwrVFIU",
"_score" : 8.61574,
"_source" : {
"id" : "c121023",
"name" : "Kemah",
"slug" : "yurtdisi/houston-ve-civari/kemah"
}
},
{
"_index" : "destinations",
"_type" : "_doc",
"_id" : "ze78Y2YBqxTX_xwrVFo5",
"_score" : 8.61574,
"_source" : {
"id" : "c1783",
"name" : "Kemerovo",
"slug" : "yurtdisi/kemerovo-oblasti/kemerovo"
}
},
{
"_index" : "destinations",
"_type" : "_doc",
"_id" : "xe78Y2YBqxTX_xwrVFs9",
"_score" : 8.61574,
"_source" : {
"id" : "c1786",
"name" : "Kemi",
"slug" : "yurtdisi/rovaniemi/kemi"
}
},
{
"_index" : "destinations",
"_type" : "_doc",
"_id" : "Tu78Y2YBqxTX_xwrVG-X",
"_score" : 8.61574,
"_source" : {
"id" : "c1900",
"name" : "Kempsey",
"slug" : "yurtdisi/new-south-wales/kempsey"
}
},
{
"_index" : "destinations",
"_type" : "_doc",
"_id" : "Bu78Y2YBqxTX_xwrVILt",
"_score" : 8.61574,
"_source" : {
"id" : "c3000010982",
"name" : "Kempley",
"slug" : "yurtdisi/dymock/kempley"
}
},
{
"_index" : "destinations",
"_type" : "_doc",
"_id" : "B-78Y2YBqxTX_xwrVILt",
"_score" : 8.61574,
"_source" : {
"id" : "c3000010983",
"name" : "Kemnay",
"slug" : "yurtdisi/inverurie/kemnay"
}
},
{
"_index" : "destinations",
"_type" : "_doc",
"_id" : "CO78Y2YBqxTX_xwrVIb_",
"_score" : 8.61574,
"_source" : {
"id" : "c3000013079",
"name" : "Kemerburgaz",
"slug" : "eyup/kemerburgaz"
}
},
{
"_index" : "destinations",
"_type" : "_doc",
"_id" : "-fL8Y2YBqxTX_xwrZQxf",
"_score" : 8.61574,
"_source" : {
"id" : "c6190744",
"name" : "Kembs",
"slug" : "yurtdisi/haut-rhin-bolge/kembs"
}
},
{
"_index" : "destinations",
"_type" : "_doc",
"_id" : "xfL8Y2YBqxTX_xwrZSG-",
"_score" : 8.61574,
"_source" : {
"id" : "c6216986",
"name" : "Kemang",
"slug" : "yurtdisi/cakarta/kemang"
}
}
]
}
}
现在他们得分相同，因为我猜每个人都有“Kem”。但是如果我做“匹配”或“匹配短语”，结果是一样的。
【问题讨论】：
-
您应该使用您发送的实际查询来扩充您的问题，这样可以让人们更容易提供提示和解决方案。
-
据我所知，默认情况下，
match查询不会匹配Kem到Kemerburgaz，除非您修改默认模糊度，然后它确实是一个模糊匹配： Fuzzy matching should not be used for scoring purposes—only to widen the net of matching terms in case there are misspellings. -
添加了索引/映射供您查看。它是如何模糊匹配的？我没有指定任何模糊性。
-
您正在使用 ngram 过滤器（顺便说一下，您最初省略了重要信息）。这基本上会为每个单词创建一个 ngram，这意味着当您匹配
Kem时，您实际上匹配的是 ngram 的Kem部分，所有以 Kem 开头的术语都将具有该部分。没有额外的得分，因为比赛是相等的。索引实际上并不知道原始单词是什么，它只知道每个索引单词有 2-15 个单词。Kem得分第一，因为它也匹配term过滤器 -
那我应该用什么？我正在使用 edge-ngram，因为我正在使用自动完成功能。在不同的例子中会发生什么变化？什么会给最接近的单词加分？
标签： elasticsearch