请问基于es数组进行query_string查询时，如何在highlight中返回最佳匹配的数据？
Elasticsearch | 作者 osmondy | 发布于2024年01月15日 | 阅读数：2297
ES版本7.6。
整体数据量有3亿，明细数据可能会更多，所以我将商标明细信息 trademarkName 放在了数组中，担心使用nested会影响到查询性能。用的是ik_smart分词器。
我的DSL查询语句如下：
整体数据量有3亿，明细数据可能会更多，所以我将商标明细信息 trademarkName 放在了数组中，担心使用nested会影响到查询性能。用的是ik_smart分词器。
我的DSL查询语句如下：
{
"query": {
"bool": {
"must": [
{
"function_score": {
"query": {
"query_string": {
"query": "(十三易)",
"fields": [
"trademarkName^8.5"
],
"type": "phrase",
"default_operator": "and",
"max_determinized_states": 10000,
"enable_position_increments": true,
"fuzziness": "AUTO",
"fuzzy_prefix_length": 0,
"fuzzy_max_expansions": 50,
"phrase_slop": 10,
"escape": false,
"auto_generate_synonyms_phrase_query": true,
"fuzzy_transpositions": true,
"boost": 1
}
},
"functions": [
{
"filter": {
"query_string": {
"query": "(十三易)",
"fields": [
"trademarkName^8.5"
],
"type": "phrase",
"default_operator": "and",
"max_determinized_states": 10000,
"enable_position_increments": true,
"fuzziness": "AUTO",
"fuzzy_prefix_length": 0,
"fuzzy_max_expansions": 50,
"phrase_slop": 0,
"escape": false,
"auto_generate_synonyms_phrase_query": true,
"fuzzy_transpositions": true,
"boost": 1
}
},
"weight": 280
},
{
"filter": {
"match_all": {
"boost": 1
}
},
"weight": 0.01,
"field_value_factor": {
"field": "entTypeScore",
"factor": 1,
"missing": 1,
"modifier": "square"
}
},
{
"filter": {
"match_all": {
"boost": 1
}
},
"weight": 15,
"field_value_factor": {
"field": "regCapUnify",
"factor": 1,
"missing": 0,
"modifier": "ln2p"
}
}
],
"score_mode": "sum",
"boost_mode": "multiply",
"max_boost": 100000,
"boost": 1
}
}
],
"filter": [
{
"bool": {
"must": [
{
"bool": {
"adjust_pure_negative": true,
"boost": 1
}
}
],
"adjust_pure_negative": true,
"boost": 1
}
}
],
"adjust_pure_negative": true,
"boost": 1
}
},
"_source": ["trademarkName"],
"highlight": {
"order": "score",
"fields": {
"trademarkName": {
"fragment_size": 80,
"number_of_fragments": 3
}
}
}
}{
"max_score" : 79977.836,
"hits" : [
{
"_index" : "company_v6",
"_type" : "_doc",
"_id" : "ab2d254d9ca99ab4a13170d7016d7a85",
"_score" : 79977.836,
"_source" : {
"trademarkName" : [
"易十三",
"十三属相",
"图形",
"名洋数字",
"智会智展",
"十三易"
]
},
"highlight" : {
"trademarkName" : [
"<em>易</em><em>十三</em>",
"<em>十三</em><em>易</em>",
"<em>十三</em>属相"
]
}
}
]
}
2 个回复
osmondy - 瞎折腾
赞同来自:
Ombres
赞同来自:
可以考虑业务代码层面后置做权重，亦或者自己实现标红逻辑