突出显示弹性搜索中的术语问题

Question

1. 我们创建了一个具有以下设置和映射的索引。

 PUT http://localhost:9200/essearch
{ 
"mappings": {
        "object": {
            "_all": {
                "enabled": false
            },
            "properties": {
               "content": {
                    "type": "text",
                    "term_vector": "with_positions_offsets",
                    "similarity": "classic",
                    "analyzer": "content_standard"
                },
                "content_phonic": {
                    "type": "text",
                    "term_vector": "with_positions_offsets",
                    "similarity": "classic",
                    "analyzer": "content_phonetic"
                },
                "content_stemming": {
                    "type": "text",
                    "term_vector": "with_positions_offsets",
                    "similarity": "classic",
                    "analyzer": "content_stemming"
                }
            }
        }
    },
 "settings": {
        "index": {
            "number_of_shards": "1",
            "similarity": {
                "default": {
                    "type": "classic"
                }
            },
            "max_result_window": "50000",
            "mapper": {
                "dynamic": "false"
            },
            "analysis": {
                "filter": {
                    "content_phonetic": {
                        "type": "phonetic",
                        "encoder": "doublemetaphone"
                    },
                    "StopWords": {
                        "type": "stop",
                        "stopwords": [
                            "after",
                            "all",
                            "under",
                            "very",
                            "well"]
                    }
                },
                "analyzer": {
                    "content_phonetic": {
                        "filter": [
                            "content_phonetic"
                        ],
                        "char_filter": [
                            "CharFilters"
                        ],
                        "type": "custom",
                        "tokenizer": "standard"
                    },
                    "content_stemming": {
                        "filter": [
                            "lowercase",
                            "porter_stem"
                        ],
                        "char_filter": [
                            "CharFilters"
                        ],
                        "type": "custom",
                        "tokenizer": "standard"
                    },
                    "content_standard": {
                        "filter": [
                            "lowercase",
                            "StopWords"
                        ],
                        "char_filter": [
                            "CharFilters"
                        ],
                        "type": "custom",
                        "tokenizer": "standard"
                    }
                },
                "char_filter": {
                    "CharFilters": {
                        "type": "mapping",
                        "mappings": [
                            ". => ' '",
                            "' => ' '",
                            "_ => ' '",
                            ": => ' '"
                        ]
                    }
                }
            },
            "number_of_replicas": "0"
        }
    }}

2：索引文档

 http://localhost:9200/essearch/object/1
{ "content" : "beginning thirty days after the anticipated COD. 
             Buyer shall be responsible for all natural gas and electrical imbalance charges.
             All prices shall be at the Reference Conditions.
             Buyer’s performance of its obligations under the ECSA with a form of guarantee in an amount. Seller shall assign its rights under said requests to Buyer.  Buyer shall have full dispatch rights subject to operational parameters  (including ramp rates. buyer said to me..."   }

3：执行Highlight查询

    http://localhost:9200/essearch/_search
 {
 "highlight": {
"pre_tags": [ "<term0 style='background-color:Lime'>", "<term1 style='background-color:Chocolate'>", "<term2 style='background-color:Pink'>"
],"post_tags": [ "</term0>", "</term1>", "</term2>" ],
"encoder": "html",
"fields": { "content": { "fragment_size": 50, "number_of_fragments": 0, "type": "fvh" } } },
"_source": false,
"query": {
"bool": {
  "must": [
    {
      "query_string": {
        "query": "(\"under said\") OR (said) OR (buyer)",
        "default_field": "content"}} ],
  "filter": [
    {
      "ids": {
        "values": [ "1" ] } } ] } } }

4：突出显示查询输出

    {
"took": 0,
"timed_out": false,
"_shards": {
    "total": 1,
    "successful": 1,
    "failed": 0
},
"hits": {
    "total": 1,
    "max_score": 0.30490398,
    "hits": [
        {
            "_index": "essearch",
            "_type": "object",
            "_id": "1",
            "_score": 0.30490398,
            "highlight": {
                "content": [
                    "beginning thirty days after the anticipated COD.
                    <term1 style='background-color:Chocolate'>Buyer</term1> 
                    shall be responsible for all natural gas and electrical imbalance charges.
                    All prices shall be at the Reference Conditions.Buyer’s performance of its obligations under the ECSA with a form of guarantee in an amount. Seller shall assign its rights under <term0 style='background-color:Lime'>said</term0> requests    to <term1 style='background-color:Chocolate'>Buyer</term1>. <term1 style='background-color:Chocolate'>Buyer</term1> shall have full dispatch rights subject to operational parameters (including ramp rates. <term1 style='background-color:Chocolate'>buyer</term1> <term0 style='background-color:Lime'>said</term0> to me..."
                ]    }  } ] } }

如果您看到我们根据提供的查询词的数量应用了 pre,post 标签。这里我们有 3 个带有 OR 运算符的术语，因此总共提供了三个 pre,post 标签。执行高亮查询后，它应该按照序列将term1标签应用于“said”术语，但 ES 将term0标签应用于“said”术语，而对于“买家”，其应用term1标签。

Answer 1

请尝试统一荧光笔而不是快速矢量荧光笔或为 FVH 定义一个 border_scanner。

您还可以使用特殊符号highlight_query来突出显示。通过突出显示查询，您可以指定不同的查询来进行搜索和突出显示。