如何在Elasticsearch中建模实体并便捷配置同义词实现多别名搜索

一、将球队建模为一等文档 #

把每个球队作为独立文档存储，核心字段包含主团队名称、同义词列表、官网链接（可选）。这样搜索任一标签时，都能直接定位到对应的球队文档。

1. 创建带同义词分析器的索引 #

先配置索引的分析规则，让搜索时自动把同义词映射到主名称：

PUT /nba_teams
{
  "settings": {
    "analysis": {
      "filter": {
        "team_synonyms": {
          "type": "synonym",
          "synonyms": [
            "mavs, dallasmavs, dallas mavericks => dallas mavericks",
            "portland trail blazers, trail blazers => trail blazers",
            "lakers, Lakers => los angeles lakers",
            "phoenix suns, Phoenix Suns => phoenix suns"
          ],
          "ignore_case": true
        }
      },
      "analyzer": {
        "team_search_analyzer": {
          "tokenizer": "standard",
          "filter": ["lowercase", "team_synonyms"]
        },
        "team_index_analyzer": {
          "tokenizer": "standard",
          "filter": ["lowercase"]
        }
      }
    }
  },
  "mappings": {
    "properties": {
      "team_name": {
        "type": "text",
        "analyzer": "team_index_analyzer",
        "search_analyzer": "team_search_analyzer"
      },
      "synonyms": {
        "type": "keyword"
      },
      "official_url": {
        "type": "keyword"
      }
    }
  }
}

• 索引时仅做小写转换，保留原始名称的多样性；
• 搜索时先统一转小写，再通过同义词过滤器把所有标签映射到主名称，确保搜任意标签都能匹配到目标文档。

2. 导入球队数据 #

按结构导入你的球队信息：

POST /nba_teams/_doc/1
{
  "team_name": "Dallas Mavericks",
  "synonyms": ["Mavs", "dallasmavs", "DALLAS MAVERICKS"],
  "official_url": "https://www.mavs.com/"
}

POST /nba_teams/_doc/2
{
  "team_name": "Trail Blazers",
  "synonyms": ["Portland Trail Blazers", "Trail Blazers"],
  "official_url": "https://www.nba.com/blazers"
}

POST /nba_teams/_doc/3
{
  "team_name": "Los Angeles Lakers",
  "synonyms": ["lakers", "Lakers"],
  "official_url": "https://www.nba.com/lakers/"
}

POST /nba_teams/_doc/4
{
  "team_name": "Phoenix Suns",
  "synonyms": ["phoenix suns", "Phoenix Suns"],
  "official_url": ""
}

二、便捷维护同义词的方案 #

1. 外部文件管理（适合中小规模） #

把同义词写到文本文件（如team_synonyms.txt）中，格式如下：

mavs, dallasmavs, dallas mavericks => dallas mavericks
portland trail blazers, trail blazers => trail blazers
lakers, Lakers => los angeles lakers
phoenix suns, Phoenix Suns => phoenix suns

将文件放到Elasticsearch的config目录下，然后更新索引配置引用该文件：

PUT /nba_teams/_settings
{
  "analysis": {
    "filter": {
      "team_synonyms": {
        "type": "synonym",
        "synonyms_path": "team_synonyms.txt",
        "ignore_case": true
      }
    }
  }
}

修改文件后，调用以下API即可生效，无需重建索引：

POST /nba_teams/_reload_search_analyzers

2. 同义词索引（适合大规模动态维护） #

如果同义词数量多、需要频繁更新，可以创建专门的同义词索引存储映射关系，再通过synonym_graph过滤器配合indices_analyzer实时读取。这种方式支持通过CRUD操作同义词文档，全程无需修改配置文件。

三、测试搜索效果 #

现在搜索任意标签，比如mavs、dallasmavs或DALLAS MAVERICKS，都会返回对应的Dallas Mavericks文档：

GET /nba_teams/_search
{
  "query": {
    "match": {
      "team_name": "mavs"
    }
  }
}

内容的提问来源于stack exchange，提问作者user352290