Milvus
Zilliz
Star44.3K预约演示开始使用
首页
  • Home
  • Docs

  • 用户指南

  • Schema 和数据字段

  • 分析仪

  • 多语言分析仪

多语言分析器Compatible with Milvus 2.5.11+

Milvus 执行文本分析时，通常会在 Collections 的整个文本字段中应用单个分析器。如果该分析器针对英语进行了优化，那么在处理其他语言（如中文、西班牙语或法语）所需的完全不同的标记化和词干规则时就会很吃力，从而导致召回率降低。例如，搜索西班牙语单词"teléfono"（意为"电话"）时，以英语为重点的分析器会被绊倒：它可能会去掉重音，不应用西班牙语特定的词干，导致相关结果被忽略。

多语言分析器解决了这个问题，它允许你在一个 Collections 中为一个文本字段配置多个分析器。这样，你就可以在一个文本字段中存储多语言文档，Milvus 会根据每个文档的相应语言规则分析文本。

限制

  • 此功能仅适用于基于 BM25 的文本检索和稀疏向量。更多信息，请参阅全文检索。

  • 单个 Collections 中的每个文档只能使用一个分析器，由其语言标识符字段值决定。

  • 性能可能会因分析器的复杂程度和文本数据的大小而有所不同。

概述

下图显示了在 Milvus 中配置和使用多语言分析器的工作流程：

Multi Language Analyzers Workflow 多语言分析器工作流程

  1. 配置多语言分析器：

    • 使用格式设置多语言分析器：<analyzer_name>: <analyzer_config>，其中每个analyzer_config 都遵循分析仪概述中所述的标准analyzer_params 配置。

    • 定义一个特殊标识符字段，用于确定每个文档的分析仪选择。

    • 配置default 分析器，用于处理未知语言。

  2. 创建 Collections：

    • 定义包含基本字段的 Schema：

      • primary_key：唯一文档标识符。

      • text_field：文本字段：存储原始文本内容。

      • identifier_字段：表示对每个文档使用哪个分析器。

      • vector_field：存储将由 BM25 函数生成的稀疏嵌入。

    • 配置 BM25 函数和索引参数。

  3. 插入带有语言标识符的数据：

    • 添加包含各种语言文本的文档，其中每个文档都包含一个标识符值，指定要使用的分析器。

    • Milvus 根据标识符字段选择适当的分析器，标识符未知的文档使用default 分析器。

  4. 使用特定语言分析器搜索：

    • 提供指定分析器名称的查询文本，Milvus 会使用指定的分析器处理查询。

    • 根据特定语言规则进行标记化，并根据相似度返回适合语言的搜索结果。

第 1 步：配置多分析器参数

multi_analyzer_params 是一个单独的 JSON 对象，它决定了 Milvus 如何为每个实体选择合适的分析器：

Python Java NodeJS Go cURL
multi_analyzer_params = {
  # Define language-specific analyzers
  # Each analyzer follows this format: <analyzer_name>: <analyzer_params>
  "analyzers": {
    "english": {"type": "english"},          # English-optimized analyzer
    "chinese": {"type": "chinese"},          # Chinese-optimized analyzer
    "default": {"tokenizer": "icu"}          # Required fallback analyzer
  },
  "by_field": "language",                    # Field determining analyzer selection
  "alias": {
    "cn": "chinese",                         # Use "cn" as shorthand for Chinese
    "en": "english"                          # Use "en" as shorthand for English
  }
}

Map<String, Object> analyzerParams = new HashMap<>();
analyzerParams.put("analyzers", new HashMap<String, Object>() {{
    put("english", new HashMap<String, Object>() {{
        put("type", "english");
    }});
    put("chinese", new HashMap<String, Object>() {{
        put("type", "chinese");
    }});
    put("default", new HashMap<String, Object>() {{
        put("tokenizer", "icu");
    }});
}});
analyzerParams.put("by_field", "language");
analyzerParams.put("alias", new HashMap<String, Object>() {{
    put("cn", "chinese");
    put("en", "english");
}});

const multi_analyzer_params = {
  // Define language-specific analyzers
  // Each analyzer follows this format: <analyzer_name>: <analyzer_params>
  "analyzers": {
    "english": {"type": "english"},          # English-optimized analyzer
    "chinese": {"type": "chinese"},          # Chinese-optimized analyzer
    "default": {"tokenizer": "icu"}          # Required fallback analyzer
  },
  "by_field": "language",                    # Field determining analyzer selection
  "alias": {
    "cn": "chinese",                         # Use "cn" as shorthand for Chinese
    "en": "english"                          # Use "en" as shorthand for English
  }
}

multiAnalyzerParams := map[string]any{
    "analyzers": map[string]any{
        "english": map[string]string{"type": "english"},
        "chinese": map[string]string{"type": "chinese"},
        "default": map[string]string{"tokenizer": "icu"},
    },
    "by_field": "language",
    "alias": map[string]string{
        "cn": "chinese",
        "en": "english",
    },
}

# restful
export multi_analyzer_params='{
  "analyzers": {
    "english": {
      "type": "english"
    },
    "chinese": {
      "type": "chinese"
    },
    "default": {
      "tokenizer": "icu"
    }
  },
  "by_field": "language",
  "alias": {
    "cn": "chinese",
    "en": "english"
  }
}'

参数

是否需要？

说明

规则

analyzers

是

列出 Milvus 可用于处理文本的每种特定语言分析器。analyzers 中的每个分析器都遵循以下格式：<analyzer_name>: <analyzer_params> 。
  • 使用标准analyzer_params 语法定义每个分析器（请参阅分析器概述）。
  • 添加一个关键字为default 的条目；只要存储在by_field 中的值与任何其他分析器名称不匹配，Milvus 就会返回到该分析器。

by_field

是

为每个文档存储 Milvus 应使用的语言（即分析器名称）的字段名称。

  • 必须是 Collections 中定义的VARCHAR 字段。

  • 每一行的值必须与analyzers 中列出的分析器名称（或别名）之一完全匹配。

  • 如果某一行的值缺失或找不到，Milvus 会自动应用default 分析器。

alias

无

为分析器创建快捷方式或替代名称，使它们更容易在代码中引用。每个分析器可以有一个或多个别名。

每个别名必须映射到现有的分析器键。

第 2 步：创建 Collections

创建支持多语言的 Collections 需要配置特定字段和索引：

添加字段

在这一步中，用四个基本字段定义 Collections Schema：

  • 主键字段(id)：Collections 中每个实体的唯一标识符。设置auto_id=True 可使 Milvus 自动生成这些 ID。

  • 语言指示符字段(language)：此 VARCHAR 字段对应于multi_analyzer_params 中指定的by_field 。它存储每个实体的语言标识符，告诉 Milvus 使用哪种分析器。

  • 文本内容字段(text)：这个 VARCHAR 字段存储要分析和搜索的实际文本数据。设置enable_analyzer=True 至关重要，因为它可以激活该字段的文本分析功能。multi_analyzer_params 配置直接连接到该字段，在文本数据和特定语言分析仪之间建立连接。

  • 向量字段(sparse)：该字段将存储 BM25 函数生成的稀疏向量。这些向量代表文本数据的可分析形式，也是 Milvus 实际搜索的内容。

Python Java NodeJS Go cURL
# Import required modules
from pymilvus import MilvusClient, DataType, Function, FunctionType

# Initialize client
client = MilvusClient(
    uri="http://localhost:19530",
)

# Initialize a new schema
schema = client.create_schema()

# Step 2.1: Add a primary key field for unique document identification
schema.add_field(
    field_name="id",                  # Field name
    datatype=DataType.INT64,          # Integer data type
    is_primary=True,                  # Designate as primary key
    auto_id=True                      # Auto-generate IDs (recommended)
)

# Step 2.2: Add language identifier field
# This MUST match the "by_field" value in language_analyzer_config
schema.add_field(
    field_name="language",       # Field name
    datatype=DataType.VARCHAR,   # String data type
    max_length=255               # Maximum length (adjust as needed)
)

# Step 2.3: Add text content field with multi-language analysis capability
schema.add_field(
    field_name="text",                           # Field name
    datatype=DataType.VARCHAR,                   # String data type
    max_length=8192,                             # Maximum length (adjust based on expected text size)
    enable_analyzer=True,                        # Enable text analysis
    multi_analyzer_params=multi_analyzer_params  # Connect with our language analyzers
)

# Step 2.4: Add sparse vector field to store the BM25 output
schema.add_field(
    field_name="sparse",                   # Field name
    datatype=DataType.SPARSE_FLOAT_VECTOR  # Sparse vector data type
)

import com.google.gson.JsonObject;
import io.milvus.common.clientenum.FunctionType;
import io.milvus.v2.client.ConnectConfig;
import io.milvus.v2.client.MilvusClientV2;
import io.milvus.v2.common.DataType;
import io.milvus.v2.common.IndexParam;
import io.milvus.v2.service.collection.request.AddFieldReq;
import io.milvus.v2.service.collection.request.CreateCollectionReq;
import io.milvus.v2.service.collection.request.DropCollectionReq;
import io.milvus.v2.service.utility.request.FlushReq;
import io.milvus.v2.service.vector.request.InsertReq;
import io.milvus.v2.service.vector.request.SearchReq;
import io.milvus.v2.service.vector.request.data.EmbeddedText;
import io.milvus.v2.service.vector.response.SearchResp;

MilvusClientV2 client = new MilvusClientV2(ConnectConfig.builder()
        .uri("http://localhost:19530")
        .build());
        
CreateCollectionReq.CollectionSchema collectionSchema = CreateCollectionReq.CollectionSchema.builder()
        .build();
        
collectionSchema.addField(AddFieldReq.builder()
        .fieldName("id")
        .dataType(DataType.Int64)
        .isPrimaryKey(true)
        .autoID(true)
        .build());
        
collectionSchema.addField(AddFieldReq.builder()
        .fieldName("language")
        .dataType(DataType.VarChar)
        .maxLength(255)
        .build());

collectionSchema.addField(AddFieldReq.builder()
        .fieldName("text")
        .dataType(DataType.VarChar)
        .maxLength(8192)
        .enableAnalyzer(true)
        .multiAnalyzerParams(analyzerParams)
        .build());
        
collectionSchema.addField(AddFieldReq.builder()
        .fieldName("sparse")
        .dataType(DataType.SparseFloatVector)
        .build());

import { MilvusClient, DataType, FunctionType } from "@zilliz/milvus2-sdk-node";

// Initialize client
const client = new MilvusClient({
  address: "http://localhost:19530",
});

// Initialize schema array
const schema = [
  {
    name: "id",
    data_type: DataType.Int64,
    is_primary_key: true,
    auto_id: true,
  },
  {
    name: "language",
    data_type: DataType.VarChar,
    max_length: 255,
  },
  {
    name: "text",
    data_type: DataType.VarChar,
    max_length: 8192,
    enable_analyzer: true,
    analyzer_params: multi_analyzer_params,
  },
  {
    name: "sparse",
    data_type: DataType.SparseFloatVector,
  },
];


import (
    "context"
    "fmt"

    "github.com/milvus-io/milvus/client/v2/column"
    "github.com/milvus-io/milvus/client/v2/entity"
    "github.com/milvus-io/milvus/client/v2/index"
    "github.com/milvus-io/milvus/client/v2/milvusclient"
)

client, err := milvusclient.New(ctx, &milvusclient.ClientConfig{
    Address: "localhost:19530",
    APIKey:  "root:Milvus",
})
if err != nil {
    fmt.Println(err.Error())
    // handle error
}

schema := entity.NewSchema()

schema.WithField(entity.NewField().
    WithName("id").
    WithDataType(entity.FieldTypeInt64).
    WithIsPrimaryKey(true).
    WithIsAutoID(true),
).WithField(entity.NewField().
    WithName("language").
    WithDataType(entity.FieldTypeVarChar).
    WithMaxLength(255),
).WithField(entity.NewField().
    WithName("text").
    WithDataType(entity.FieldTypeVarChar).
    WithMaxLength(8192).
    WithEnableAnalyzer(true).
    WithMultiAnalyzerParams(multiAnalyzerParams),
).WithField(entity.NewField().
    WithName("sparse").
    WithDataType(entity.FieldTypeSparseVector),
)

# restful
export TOKEN="root:Milvus"
export CLUSTER_ENDPOINT="http://localhost:19530"

export idField='{
  "fieldName": "id",
  "dataType": "Int64",
  "isPrimary": true,
  "autoID": true
}'

export languageField='{
  "fieldName": "language",
  "dataType": "VarChar",
  "elementTypeParams": {
    "max_length": 255
  }
}'

export textField='{
  "fieldName": "text",
  "dataType": "VarChar",
  "elementTypeParams": {
    "max_length": 8192,
    "enable_analyzer": true，
    "multiAnalyzerParam": '"$multi_analyzer_params"'
  },
}'

export sparseField='{
  "fieldName": "sparse",
  "dataType": "SparseFloatVector"
}'

定义 BM25 函数

定义一个 BM25 函数，以便从原始文本数据中生成稀疏向量表示：

Python Java NodeJS Go cURL
# Create the BM25 function
bm25_function = Function(
    name="text_to_vector",            # Descriptive function name
    function_type=FunctionType.BM25,  # Use BM25 algorithm
    input_field_names=["text"],       # Process text from this field
    output_field_names=["sparse"]     # Store vectors in this field
)

# Add the function to our schema
schema.add_function(bm25_function)

CreateCollectionReq.Function function = CreateCollectionReq.Function.builder()
        .functionType(FunctionType.BM25)
        .name("text_to_vector")
        .inputFieldNames(Collections.singletonList("text"))
        .outputFieldNames(Collections.singletonList("sparse"))
        .build();
collectionSchema.addFunction(function);

const functions = [
  {
    name: "text_bm25_emb",
    description: "bm25 function",
    type: FunctionType.BM25,
    input_field_names: ["text"],
    output_field_names: ["sparse"],
    params: {},
  },
];

function := entity.NewFunction()
schema.WithFunction(function.WithName("text_to_vector").
    WithType(entity.FunctionTypeBM25).
    WithInputFields("text").
    WithOutputFields("sparse"))

# restful
export function='{
  "name": "text_to_vector",
  "type": "BM25",
  "inputFieldNames": ["text"],
  "outputFieldNames": ["sparse"]
}'

export schema="{
  \"autoID\": true,
  \"fields\": [
    $idField,
    $languageField,
    $textField,
    $sparseField
  ],
  \"functions\": [
    $function
  ]
}"

该函数会根据每个文本条目的语言标识符自动应用相应的分析器。有关基于 BM25 的文本检索的更多信息，请参阅全文检索。

配置索引参数

为实现高效搜索，请在稀疏向量场上创建索引：

Python Java NodeJS Go cURL
# Configure index parameters
index_params = client.prepare_index_params()

# Add index for sparse vector field
index_params.add_index(
    field_name="sparse",        # Field to index (our vector field)
    index_type="AUTOINDEX",     # Let Milvus choose optimal index type
    metric_type="BM25"          # Must be BM25 for this feature
)

List<IndexParam> indexes = new ArrayList<>();
indexes.add(IndexParam.builder()
        .fieldName("sparse")
        .indexType(IndexParam.IndexType.AUTOINDEX)
        .metricType(IndexParam.MetricType.BM25)
        .build());

const index_params = [{
    field_name: "sparse",
    index_type: "AUTOINDEX",
    metric_type: "BM25"
}];

idx := index.NewAutoIndex(index.MetricType(entity.BM25))
indexOption := milvusclient.NewCreateIndexOption("multilingual_documents", "sparse", idx)

# restful
export IndexParams='[
  {
    "fieldName": "sparse",
    "indexType": "AUTOINDEX",
    "metricType": "BM25",
    "params": {}
  }
]'

索引通过组织稀疏向量来提高搜索性能，从而实现高效的 BM25 相似性计算。

创建 Collections

最后的创建步骤将您之前的所有配置汇集在一起：

  • collection_name="multilang_demo" 为你的 Collection 命名，以备将来参考。

  • schema=schema 应用您定义的字段结构和功能。

  • index_params=index_params 实施索引策略，实现高效搜索。

Python Java NodeJS Go cURL
# Create collection
COLLECTION_NAME = "multilingual_documents"

# Check if collection already exists
if client.has_collection(COLLECTION_NAME):
    client.drop_collection(COLLECTION_NAME)  # Remove it for this example
    print(f"Dropped existing collection: {COLLECTION_NAME}")

# Create the collection
client.create_collection(
    collection_name=COLLECTION_NAME,       # Collection name
    schema=schema,                         # Our multilingual schema
    index_params=index_params              # Our search index configuration
)

client.dropCollection(DropCollectionReq.builder()
        .collectionName("multilingual_documents")
        .build());
        
CreateCollectionReq requestCreate = CreateCollectionReq.builder()
        .collectionName("multilingual_documents")
        .collectionSchema(collectionSchema)
        .indexParams(indexes)
        .build();
client.createCollection(requestCreate);

const COLLECTION_NAME = "multilingual_documents";

// Create the collection
await client.createCollection({
  collection_name: COLLECTION_NAME,
  schema: schema,
  index_params: index_params,
  functions: functions
});


err = client.CreateCollection(ctx,
    milvusclient.NewCreateCollectionOption("multilingual_documents", schema).
        WithIndexOptions(indexOption))
if err != nil {
    fmt.Println(err.Error())
    // handle error
}

# restful
curl --request POST \
--url "${CLUSTER_ENDPOINT}/v2/vectordb/collections/create" \
--header "Authorization: Bearer ${TOKEN}" \
--header "Content-Type: application/json" \
--data "{
  \"collectionName\": \"multilingual_documents\",
  \"schema\": $schema,
  \"indexParams\": $IndexParams
}"

此时，Milvus 会创建一个支持多语言分析器的空 Collections，随时准备接收数据。

第 3 步：插入示例数据

向多语言 Collections 添加文档时，每个文档都必须包含文本内容和语言标识符：

Python Java NodeJS Go cURL
# Prepare multilingual documents
documents = [
    # English documents
    {
        "text": "Artificial intelligence is transforming technology",
        "language": "english",  # Using full language name
    },
    {
        "text": "Machine learning models require large datasets",
        "language": "en",  # Using our defined alias
    },
    # Chinese documents
    {
        "text": "人工智能正在改变技术领域",
        "language": "chinese",  # Using full language name
    },
    {
        "text": "机器学习模型需要大型数据集",
        "language": "cn",  # Using our defined alias
    },
]

# Insert the documents
result = client.insert(COLLECTION_NAME, documents)

# Print results
inserted = result["insert_count"]            
print(f"Successfully inserted {inserted} documents")
print("Documents by language: 2 English, 2 Chinese")

# Expected output:
# Successfully inserted 4 documents
# Documents by language: 2 English, 2 Chinese

List<String> texts = Arrays.asList(
        "Artificial intelligence is transforming technology",
        "Machine learning models require large datasets",
        "人工智能正在改变技术领域",
        "机器学习模型需要大型数据集"
);
List<String> languages = Arrays.asList(
        "english", "en", "chinese", "cn"
);

List<JsonObject> rows = new ArrayList<>();
for (int i = 0; i < texts.size(); i++) {
    JsonObject row = new JsonObject();
    row.addProperty("text", texts.get(i));
    row.addProperty("language", languages.get(i));
    rows.add(row);
}
client.insert(InsertReq.builder()
        .collectionName("multilingual_documents")
        .data(rows)
        .build());

// Prepare multilingual documents
const documents = [
  // English documents
  {
    text: "Artificial intelligence is transforming technology",
    language: "english",
  },
  {
    text: "Machine learning models require large datasets",
    language: "en",
  },
  // Chinese documents
  {
    text: "人工智能正在改变技术领域",
    language: "chinese",
  },
  {
    text: "机器学习模型需要大型数据集",
    language: "cn",
  },
];

// Insert the documents
const result = await client.insert({
  collection_name: COLLECTION_NAME,
  data: documents,
});

// Print results
const inserted = result.insert_count;
console.log(`Successfully inserted ${inserted} documents`);
console.log("Documents by language: 2 English, 2 Chinese");

// Expected output:
// Successfully inserted 4 documents
// Documents by language: 2 English, 2 Chinese


column1 := column.NewColumnVarChar("text",
    []string{
        "Artificial intelligence is transforming technology",
        "Machine learning models require large datasets",
        "人工智能正在改变技术领域",
        "机器学习模型需要大型数据集",
    })
column2 := column.NewColumnVarChar("language",
    []string{"english", "en", "chinese", "cn"})

_, err = client.Insert(ctx, milvusclient.NewColumnBasedInsertOption("multilingual_documents").
    WithColumns(column1, column2),
)
if err != nil {
    fmt.Println(err.Error())
    // handle err
}

# restful
curl --request POST \
--url "${CLUSTER_ENDPOINT}/v2/vectordb/entities/insert" \
--header "Authorization: Bearer ${TOKEN}" \
--header "Content-Type: application/json" \
--data '{
  "collectionName": "multilingual_documents",
  "data": [
    {
      "text": "Artificial intelligence is transforming technology",
      "language": "english"
    },
    {
      "text": "Machine learning models require large datasets",
      "language": "en"
    },
    {
      "text": "人工智能正在改变技术领域",
      "language": "chinese"
    },
    {
      "text": "机器学习模型需要大型数据集",
      "language": "cn"
    }
  ]
}'

在插入过程中，Milvus

  1. 读取每个文档的language 字段

  2. 对text 字段应用相应的分析器

  3. 通过 BM25 函数生成稀疏向量表示法

  4. 存储原始文本和生成的稀疏向量

您无需直接提供稀疏向量；BM25 函数会根据您的文本和指定的分析器自动生成稀疏向量。

步骤 4：执行搜索操作符

使用英文分析器

使用多语言分析器搜索时，search_params 包含关键配置：

  • metric_type="BM25" 必须与您的索引配置相匹配。

  • analyzer_name="english" 指定对查询文本应用哪种分析器。这与存储文档中使用的分析器无关。

  • params={"drop_ratio_search": "0"} 控制特定于 BM25 的行为；在这里，它会保留搜索中的所有术语。更多信息，请参阅稀疏向量。

Python Java NodeJS Go cURL
search_params = {
    "metric_type": "BM25",            # Must match index configuration
    "analyzer_name": "english",  # Analyzer that matches the query language
    "drop_ratio_search": "0",     # Keep all terms in search (tweak as needed)
}

# Execute the search
english_results = client.search(
    collection_name=COLLECTION_NAME,  # Collection to search
    data=["artificial intelligence"],                # Query text
    anns_field="sparse",              # Field to search against
    search_params=search_params,      # Search configuration
    limit=3,                      # Max results to return
    output_fields=["text", "language"],  # Fields to include in the output
    consistency_level="Bounded",       # Data‑consistency guarantee
)

# Display English search results
print("\n=== English Search Results ===")
for i, hit in enumerate(english_results[0]):
    print(f"{i+1}. [{hit.score:.4f}] {hit.entity.get('text')} "
          f"(Language: {hit.entity.get('language')})")

# Expected output:
# === English Search Results ===
# 1. [2.7881] Artificial intelligence is transforming technology (Language: english)

Map<String,Object> searchParams = new HashMap<>();
searchParams.put("metric_type", "BM25");
searchParams.put("analyzer_name", "english");
searchParams.put("drop_ratio_search", 0);
SearchResp searchResp = client.search(SearchReq.builder()
        .collectionName("multilingual_documents")
        .data(Collections.singletonList(new EmbeddedText("artificial intelligence")))
        .annsField("sparse")
        .topK(3)
        .searchParams(searchParams)
        .outputFields(Arrays.asList("text", "language"))
        .build());

System.out.println("\n=== English Search Results ===");
List<List<SearchResp.SearchResult>> searchResults = searchResp.getSearchResults();
for (List<SearchResp.SearchResult> results : searchResults) {
    for (SearchResp.SearchResult result : results) {
        System.out.printf("Score: %f, %s\n", result.getScore(), result.getEntity().toString());
    }
}

// Execute the search
const english_results = await client.search({
  collection_name: COLLECTION_NAME,
  data: ["artificial intelligence"],
  anns_field: "sparse",
  params: {
    metric_type: "BM25",
    analyzer_name: "english",
    drop_ratio_search: "0",
  },
  limit: 3,
  output_fields: ["text", "language"],
  consistency_level: "Bounded",
});

// Display English search results
console.log("\n=== English Search Results ===");
english_results.results.forEach((hit, i) => {
  console.log(
    `${i + 1}. [${hit.score.toFixed(4)}] ${hit.entity.text} ` +
      `(Language: ${hit.entity.language})`
  );
});


annSearchParams := index.NewCustomAnnParam()
annSearchParams.WithExtraParam("metric_type", "BM25")
annSearchParams.WithExtraParam("analyzer_name", "english")
annSearchParams.WithExtraParam("drop_ratio_search", 0)

resultSets, err := client.Search(ctx, milvusclient.NewSearchOption(
    "multilingual_documents", // collectionName
    3,                        // limit
    []entity.Vector{entity.Text("artificial intelligence")},
).WithANNSField("sparse").
    WithAnnParam(annSearchParams).
    WithOutputFields("text", "language"))
if err != nil {
    fmt.Println(err.Error())
    // handle error
}

for _, resultSet := range resultSets {
    for i := 0; i < len(resultSet.Scores); i++ {
        text, _ := resultSet.GetColumn("text").GetAsString(i)
        lang, _ := resultSet.GetColumn("language").GetAsString(i)
        fmt.Println("Score: ", resultSet.Scores[i], "Text: ", text, "Language:", lang)
    }
}

# restful
curl --request POST \
--url "${CLUSTER_ENDPOINT}/v2/vectordb/entities/search" \
--header "Authorization: Bearer ${TOKEN}" \
--header "Content-Type: application/json" \
--data '{
  "collectionName": "multilingual_documents",
  "data": ["artificial intelligence"],
  "annsField": "sparse",
  "limit": 3,
  "searchParams": {
    "metric_type": "BM25",
    "analyzer_name": "english",
    "drop_ratio_search": "0"  
  },
  "outputFields": ["text", "language"],
  "consistencyLevel": "Bounded"
}'

使用中文分析器

本示例演示了针对不同的查询文本切换到中文分析器（使用其别名"cn" ）。所有其他参数保持不变，但现在使用特定于中文的标记化规则处理查询文本。

Python Java NodeJS Go cURL
search_params["analyzer_name"] = "cn"

chinese_results = client.search(
    collection_name=COLLECTION_NAME,  # Collection to search
    data=["人工智能"],                # Query text
    anns_field="sparse",              # Field to search against
    search_params=search_params,      # Search configuration
    limit=3,                      # Max results to return
    output_fields=["text", "language"],  # Fields to include in the output
    consistency_level="Bounded",       # Data‑consistency guarantee
)

# Display Chinese search results
print("\n=== Chinese Search Results ===")
for i, hit in enumerate(chinese_results[0]):
    print(f"{i+1}. [{hit.score:.4f}] {hit.entity.get('text')} "
          f"(Language: {hit.entity.get('language')})")

# Expected output:
# === Chinese Search Results ===
# 1. [3.3814] 人工智能正在改变技术领域 (Language: chinese)

searchParams.put("analyzer_name", "cn");
searchResp = client.search(SearchReq.builder()
        .collectionName("multilingual_documents")
        .data(Collections.singletonList(new EmbeddedText("人工智能")))
        .annsField("sparse")
        .topK(3)
        .searchParams(searchParams)
        .outputFields(Arrays.asList("text", "language"))
        .build());

System.out.println("\n=== Chinese Search Results ===");
searchResults = searchResp.getSearchResults();
for (List<SearchResp.SearchResult> results : searchResults) {
    for (SearchResp.SearchResult result : results) {
        System.out.printf("Score: %f, %s\n", result.getScore(), result.getEntity().toString());
    }
}

// Execute the search
const cn_results = await client.search({
  collection_name: COLLECTION_NAME,
  data: ["人工智能"],
  anns_field: "sparse",
  params: {
    metric_type: "BM25",
    analyzer_name: "cn",
    drop_ratio_search: "0",
  },
  limit: 3,
  output_fields: ["text", "language"],
  consistency_level: "Bounded",
});

// Display Chinese search results
console.log("\n=== Chinese Search Results ===");
cn_results.results.forEach((hit, i) => {
  console.log(
    `${i + 1}. [${hit.score.toFixed(4)}] ${hit.entity.text} ` +
      `(Language: ${hit.entity.language})`
  );
});


annSearchParams.WithExtraParam("analyzer_name", "cn")

resultSets, err = client.Search(ctx, milvusclient.NewSearchOption(
    "multilingual_documents", // collectionName
    3,                        // limit
    []entity.Vector{entity.Text("人工智能")},
).WithANNSField("sparse").
    WithAnnParam(annSearchParams).
    WithOutputFields("text", "language"))
if err != nil {
    fmt.Println(err.Error())
    // handle error
}

for _, resultSet := range resultSets {
    for i := 0; i < len(resultSet.Scores); i++ {
        text, _ := resultSet.GetColumn("text").GetAsString(i)
        lang, _ := resultSet.GetColumn("language").GetAsString(i)
        fmt.Println("Score: ", resultSet.Scores[i], "Text: ", text, "Language:", lang)
    }
}


# restful
curl --request POST \
--url "${CLUSTER_ENDPOINT}/v2/vectordb/entities/search" \
--header "Authorization: Bearer ${TOKEN}" \
--header "Content-Type: application/json" \
--data '{
  "collectionName": "multilingual_documents",
  "data": ["人工智能"],
  "annsField": "sparse",
  "limit": 3,
  "searchParams": {
    "analyzer_name": "cn"
  },
  "outputFields": ["text", "language"],
  "consistencyLevel": "Bounded"
}'
目录

想要更快、更简单、更好用的 Milvus SaaS服务 ？

Zilliz Cloud是基于Milvus的全托管向量数据库，拥有更高性能，更易扩展，以及卓越性价比

免费试用 Zilliz Cloud
反馈

此页对您是否有帮助?