求助 SpringAI 中的 similaritySearch 用法

我自己用Redis数据库实现了向量的存储，但是在进行相似度检索的时候，一旦加入了过滤条件，如eq,ne等等之类的，就获取不到任何文档。

首先maven导入相关的依赖如下：

`

▼
text
复制代码
   <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-model-deepseek</artifactId>
        <version>1.0.0</version>
    </dependency>

    <dependency>
        <groupId>com.alibaba.cloud.ai</groupId>
        <artifactId>spring-ai-alibaba-starter-dashscope</artifactId>
        <version>1.0.0.2</version>
    </dependency>

    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-redis-store</artifactId>
         <version>1.0.0</version>
    </dependency>


    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-autoconfigure-vector-store-redis</artifactId>
         <version>1.0.0</version>
    </dependency>

    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-starter-vector-store-redis</artifactId>
         <version>1.0.0</version>
    </dependency>

    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-advisors-vector-store</artifactId>
         <version>1.0.0</version>
    </dependency>


    <!-- 强制所有模块使用 tika-parser-pdf-module 3.1.0 ，兼容SpringAI的TIKA-->
    <dependency>
        <groupId>org.apache.tika</groupId>
        <artifactId>tika-parser-pdf-module</artifactId>
        <version>3.1.0</version>
    </dependency>
    <!-- 也强制使用 tika-core 配合 -->
    <dependency>
        <groupId>org.apache.tika</groupId>
        <artifactId>tika-core</artifactId>
        <version>3.1.0</version>
    </dependency>

    <dependency>
        <groupId>org.apache.tika</groupId>
        <artifactId>tika-parser-ocr-module</artifactId>
        <version>3.1.0</version>
    </dependency>

    <dependency>
        <groupId>org.apache.pdfbox</groupId>
        <artifactId>pdfbox</artifactId>
        <version>2.0.29</version>
    </dependency>

    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-pdf-document-reader</artifactId>
         <version>1.0.0</version>
    </dependency>

    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-tika-document-reader</artifactId>
         <version>1.0.0</version>
        <exclusions>
            <exclusion>
                <groupId>org.apache.tika</groupId>
                <artifactId>tika-parser-pdf-module</artifactId>
            </exclusion>
            <exclusion>
                <groupId>org.apache.tika</groupId>
                <artifactId>tika-core</artifactId>
            </exclusion>
        </exclusions>
    </dependency>

    <dependency>
        <groupId>org.springframework.ai</groupId>
        <artifactId>spring-ai-markdown-document-reader</artifactId>
         <version>1.0.0</version>
    </dependency>

`

相关的配置类如下

▼
text
复制代码
@Configuration
public class AICourseAppVectorStoreConfig {

    @Resource
    private AICourseDocumentLoader aiCourseDocumentLoader;

    @Resource(name = "dashscopeEmbeddingModel")
    private EmbeddingModel dashscopeEmbeddingModel;

    @Resource
    private MyTokenTextSplitter myTokenTextSplitter;


    @Bean
    public SimpleVectorStore simpleVectorStore() {
        //这里的AI嵌入模型使用的是springAI官方的模型
        SimpleVectorStore simpleVectorStore = SimpleVectorStore.builder(dashscopeEmbeddingModel).build();
        return simpleVectorStore;
    }


    @Bean
    public RedisVectorStoreProperties redisVectorStoreProperties() {
        RedisVectorStoreProperties properties = new RedisVectorStoreProperties();
        properties.setIndexName("Intelli-Edu-index");
        properties.setPrefix("Intelli-Edu-prefix");
        properties.setInitializeSchema(true);
        return properties;
    }


    @Bean
    public RedisVectorStore redisVectorStore(
            @Value("${spring.data.redis.host}") String redisHost,
            @Value("${spring.data.redis.port}") int redisPort,
            @Value("${spring.data.redis.password}") String redisPassword,
            @Value("${spring.data.redis.database}") int redisDatabase,
            RedisVectorStoreProperties properties
    ) {
        JedisClientConfig clientConfig = DefaultJedisClientConfig.builder()
                .password(redisPassword)
                .database(redisDatabase)
                .build();

        JedisPooled jedisPooled = new JedisPooled(new HostAndPort(redisHost, redisPort), clientConfig);

        RedisVectorStore redisVectorStore =
                RedisVectorStore.builder(jedisPooled, dashscopeEmbeddingModel)
                .metadataFields(Arrays.asList(RedisVectorStore.MetadataField.tag("tag"),
                        RedisVectorStore.MetadataField.tag("accessPermission"),
                        RedisVectorStore.MetadataField.tag("filename")

                ))
                .indexName(properties.getIndexName())
                .prefix(properties.getPrefix())
                .initializeSchema(properties.isInitializeSchema())
                .build();
        return redisVectorStore;
    }
}
`

扩展了CustomTikaDocumentReader(原生的TikaDocumentReader不支持自定义元数据)如下

▼
text
复制代码
public class CustomTikaDocumentReader extends TikaDocumentReader {

    private final Map<String,Object> customMetadata;

    public CustomTikaDocumentReader(Resource resource, Map<String, Object> customMetadata) {
        super(resource);
        this.customMetadata = customMetadata;
    }


    @Override
    public List<Document> get() {
        List<Document> documents = super.get();
        for (Document document : documents) {
            // 将自定义元数据添加到每个文档中
            if (Objects.nonNull(document.getMetadata()) ) {
                customMetadata.forEach(document.getMetadata()::put);
            }
        }
        return documents;
    }

}

插入向量数据的数据如图所示

▼
text
复制代码
{
  "filename": "智慧的疆界.md",
  "accessPermission": "public",
  "embedding": [
  -0.45072716,
    -2.7041578,
    -5.0076756,
    -0.6626291,
         .
         .
         .
      ],
  "tag": "AI通识历史;深度学习;人工智能",
  "category": "header_1",
  "title": "7.5.3 深度学习时代",
  "content": "2006年被认为深度学习时代的元年，省略多个字~",
  "status": "AI通识历史"
  }

官方的检索示例如图所示

自己的查询代码如下：

▼
text
复制代码
    Filter.Expression expression = new FilterExpressionBuilder()
            .in("accessPermission", "public")

            .build();
    log.info("搜索条件: {}", expression.toString());

    List<Document> results =  redisVectorStore.similaritySearch(SearchRequest.builder()
            .query("介绍下香农")
            .topK(5)
            .similarityThreshold(0.7)
            .filterExpression(expression)
            .build());
    log.info("搜索结果: {}", results);

检索结果为空。去掉Filter.Expression 的过滤条件则能够正常返回结果如下图所示。

请各位解答