我在一个分片上有一个 Elasticsearch 索引，我正在为此运行一个[dis_max][4]查询，给定一些用户详细信息
（名字、姓氏、出生日期、地址、电话、用户名、电子邮件等）
从组合了一组标准/匹配子句的索引中查询用户。
例如
- 匹配用户名(
[fuzzy][1], boosted 2x) - 应该匹配名字和姓氏（
[bool][3]结合[match-term][2]查询 FN 和 LN，提升 1.1 倍） - 必须匹配 FN、LN 和 DOB（
[bool][3]结合[fuzzy][1]FN 和 LN 以及[match-term][2]DOB，提升 3 倍） - 匹配手机（
[match-term][2]提升 2 倍）
等等
请参阅下面的查询（输入数据模糊）：
{
"from":0,
"size":100,
"explain": true,
"query":{
"dis_max":{
"tie_breaker":0.5,
"queries":[
{
"fuzzy":{
"username":{
"value":"xxx",
"fuzziness":"AUTO",
"prefix_length":0,
"max_expansions":50,
"transpositions":false,
"boost":2.0
}
}
},
{
"term":{
"email":{
"value":"gmail.com",
"boost":1.0
}
}
},
{
"fuzzy":{
"email":{
"value":"xxx",
"fuzziness":"AUTO",
"prefix_length":0,
"max_expansions":50,
"transpositions":false,
"boost":1.0
}
}
},
{
"term":{
"password-hash":{
"value":"xxx",
"boost":1.0
}
}
},
{
"term":{
"currency-code":{
"value":"xxx",
"boost":0.5
}
}
},
{
"match":{
"first-name":{
"query":"xxx",
"operator":"OR",
"prefix_length":0,
"max_expansions":50,
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.0
}
}
},
{
"match":{
"last-name":{
"query":"xxx",
"operator":"OR",
"prefix_length":0,
"max_expansions":50,
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.0
}
}
},
{
"dis_max":{
"queries":[
{
"match":{
"first-name":{
"query":"xxx",
"operator":"OR",
"prefix_length":0,
"max_expansions":50,
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.0
}
}
},
{
"match":{
"last-name":{
"query":"xxx",
"operator":"OR",
"prefix_length":0,
"max_expansions":50,
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.0
}
}
}
],
"boost":1.1
}
},
{
"match":{
"date-of-birth":{
"query":"xxxx-xx-xx",
"operator":"OR",
"prefix_length":0,
"max_expansions":50,
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.0
}
}
},
{
"bool":{
"must":[
{
"fuzzy" : {
"first-name" : {
"value" : "xxx",
"fuzziness" : "AUTO",
"prefix_length" : 0,
"max_expansions" : 50,
"transpositions" : true,
"boost" : 1.5
}
}
},
{
"fuzzy" : {
"last-name" : {
"value" : "xxx",
"fuzziness" : "AUTO",
"prefix_length" : 0,
"max_expansions" : 50,
"transpositions" : true,
"boost" : 1.5
}
}
},
{
"match":{
"date-of-birth":{
"query":"xxxx-xx-xx",
"operator":"OR",
"prefix_length":0,
"max_expansions":50,
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.0
}
}
}
],
"adjust_pure_negative":true,
"boost":3.0
}
},
{
"match":{
"address":{
"query":"xxx",
"operator":"OR",
"analyzer":"whitespace",
"fuzziness":"AUTO",
"prefix_length":0,
"max_expansions":50,
"minimum_should_match":"60%",
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.1
}
}
},
{
"match":{
"city":{
"query":"xxx",
"operator":"OR",
"prefix_length":0,
"max_expansions":50,
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.0
}
}
},
{
"term":{
"postal-code":{
"value":"xxx ",
"boost":1.5
}
}
},
{
"boosting":{
"positive":{
"match":{
"address":{
"query":"xxx",
"operator":"OR",
"analyzer":"whitespace",
"fuzziness":"AUTO",
"prefix_length":0,
"max_expansions":50,
"minimum_should_match":"60%",
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.1
}
}
},
"negative":{
"bool":{
"must_not":[
{
"term":{
"postal-code":{
"value":"xxx ",
"boost":1.5
}
}
}
],
"adjust_pure_negative":true,
"boost":1.0
}
},
"negative_boost":0.7,
"boost":1.0
}
},
{
"term":{
"country-code":{
"value":"xxx",
"boost":1.0
}
}
},
{
"match":{
"phone":{
"query":"xxx",
"operator":"OR",
"prefix_length":0,
"max_expansions":50,
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":2.0
}
}
},
{
"match":{
"security-question":{
"query":"xxx?",
"operator":"OR",
"prefix_length":0,
"max_expansions":50,
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.0
}
}
},
{
"match":{
"security-answer":{
"query":"xxx ",
"operator":"OR",
"prefix_length":0,
"max_expansions":50,
"fuzzy_transpositions":true,
"lenient":false,
"zero_terms_query":"NONE",
"auto_generate_synonyms_phrase_query":true,
"boost":1.0
}
}
}
],
"boost":1.0
}
}
}所有标准都考虑一个分数
，我设置了一个tie_breaker，0.5以便结果的分数将是所有分数中的最大值，再加0.5上其余分数的倍数。
使用很少的输入组合执行查询，
- 在某些情况下，我得到了很好的分数，可以很好地匹配，
- 在其他情况下，即使期望分数相同或足够高，我也会得到非常低的分数，因为似乎跳过了一些最相关的匹配子句。
"explain": true事实上，我已经在解释中调试了查询执行
- 第一个结果在所有查询子句中得分很高，
- 第二个（从数据中得分应该足够）只是得分较低，并且某些子句没有出现在解释中，就好像它们被排除/忽略了一样。
我想了解为什么在某些情况下会忽略/跳过这些。有人知道这是否可能是 ES 在 Solr 中构建查询的方式中的一个问题？
请参见下面的结果示例（所有数据都被遮盖了，但结果在不同的字段中会非常接近）。
{
"took": 312,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 10000,
"relation": "gte"
},
"max_score": 164.04868,
"hits": [
{
"_shard": "[my-shard][0]",
"_node": "acccSjFfQnOffqbiHV5nAg",
"_index": "my-index",
"_type": "_doc",
"_id": "360086",
"_score": 164.04868,
"_source": {
"id": "360086",
"user-id": 389872,
"username": "xxx",
"email": "xxx@gmail.com",
"password-hash": "xxx",
"currency-code": "xxx",
"first-name": "xxx",
"last-name": "xxx",
"date-of-birth": "xxxx-xx-xx",
"address": "xxx",
"city": "N/A",
"postal-code": "xxx",
"country-code": "xxx",
"phone": "xxx",
"security-question": "xxx?",
"security-answer": "xxx "
},
"_explanation": {
"value": 164.04868,
"description": "max plus 0.5 times others of:",
"details": [
{
"value": 22.275639,
"description": "weight(username:xxx in 3223) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 0.88973737,
"description": "weight(email:gmail.com in 3223) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 12.44133,
"description": "weight(email:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 11.066888,
"description": "weight(password-hash:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 12.547058,
"description": "weight(first-name:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 12.619294,
"description": "weight(last-name:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 13.881224,
"description": "max of:",
"details": [
{
"value": 13.801764,
"description": "weight(first-name:xxx in 3223) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 13.881224,
"description": "weight(last-name:xxx in 3223) [PerFieldSimilarity], result of:",
"details": [...]
}
]
},
{
"value": 1.0,
"description": "date-of-birth:[XXX TO XXX]",
"details": []
},
{
"value": 92.82605,
"description": "sum of:",
"details": [
{
"value": 46.42945,
"description": "sum of:",
"details": [
{
"value": 46.42945,
"description": "weight(first-name:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
}
]
},
{
"value": 43.3966,
"description": "sum of:",
"details": [
{
"value": 43.3966,
"description": "weight(last-name:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
}
]
},
{
"value": 3.0,
"description": "date-of-birth:[XXX TO XXX]^3.0",
"details": []
}
]
},
{
"value": 17.182709,
"description": "weight(postal-code:XXX in 3223) [PerFieldSimilarity], result of:",
"details": []
},
{
"value": 0.6058445,
"description": "weight(country-code:XXX in 3223) [PerFieldSimilarity], result of:",
"details": []
},
{
"value": 24.692732,
"description": "weight(phone:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 4.2287235,
"description": "sum of:",
"details": [...]
},
{
"value": 0.7475863,
"description": "weight(security-question:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 1.7405679,
"description": "weight(security-question:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 1.7405679,
"description": "weight(security-question:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
}
]
},
{
"value": 9.014062,
"description": "weight(security-answer:XXX in 3223) [PerFieldSimilarity], result of:",
"details": [...]
}
]
}
},
{
"_shard": "[my-shard][0]",
"_node": "acccSjFfQnOffqbiHV5nAg",
"_index": "my-index",
"_type": "_doc",
"_id": "359895",
"_score": 40.7084,
"_source": {
"id": "359895",
"user-id": 389681,
"username": "XXX",
"email": "XXX@icloud.com",
"password-hash": "XXX",
"currency-code": "XXX",
"first-name": "XXX",
"last-name": "XXX",
"date-of-birth": "1973-03-01",
"address": "XXX",
"city": "N/A",
"postal-code": "XXX ",
"country-code": "XXX",
"phone": "XXX",
"security-question": "XXX?",
"security-answer": "XXX"
},
"_explanation": {
"value": 40.7084,
"description": "max plus 0.5 times others of:",
"details": [
{
"value": 1.0,
"description": "date-of-birth:[XXX TO XXX]",
"details": []
},
{
"value": 17.182709,
"description": "weight(postal-code:XXX in 3183) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 0.6058445,
"description": "weight(country-code:XXX in 3183) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 24.692732,
"description": "weight(phone:XXX in 3183) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 4.2287235,
"description": "sum of:",
"details": [
{
"value": 1.5324217E-6,
"description": "weight(security-question:XXX in 3183) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 0.7475863,
"description": "weight(security-question:XXX in 3183) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 1.7405679,
"description": "weight(security-question:XXX in 3183) [PerFieldSimilarity], result of:",
"details": [...]
},
{
"value": 1.7405679,
"description": "weight(security-question:XXX in 3183) [PerFieldSimilarity], result of:",
"details": [...]
}
]
},
{
"value": 9.014062,
"description": "weight(security-answer:XXX in 3183) [PerFieldSimilarity], result of:",
"details": [...]
}
]
}
}