Complete Guide to Building a Spring AI + Ollama Embedding Vectorization Project

Overview

The article shows how to extend a Spring Boot + LangChain4j + Ollama RAG project with text‑embedding (vectorization) capabilities, enabling private document search and similarity analysis.

1. Application configuration

server:
  port: 885
logging:
  level:
    com.badao: debug
    org.springframework.ai: debug
spring:
  ai:
    ollama:
      base-url: http://localhost:11434
      chat:
        options:
          model: qwen2.5
          temperature: 0.7
      embedding:
        options:
          model: nomic-embed-text

Before running the service, pull the embedding model in Ollama (e.g., ollama pull nomic-embed-text) or replace it with another model such as bge-m3 or mxbai-embed-large.

2. EmbeddingService

package com.badao.ai.service;

import org.springframework.ai.embedding.EmbeddingModel;
import org.springframework.ai.embedding.EmbeddingResponse;
import org.springframework.stereotype.Service;
import java.util.*;

@Service
public class EmbeddingService {
    private final EmbeddingModel embeddingModel;
    public EmbeddingService(EmbeddingModel embeddingModel) { this.embeddingModel = embeddingModel; }

    /** Convert a single text to a float vector */
    public float[] embed(String text) { return embeddingModel.embed(text); }

    /** Batch conversion returning the full response */
    public EmbeddingResponse embedForResponse(List<String> texts) { return embeddingModel.embedForResponse(texts); }

    /** Cosine similarity between two texts */
    public double cosineSimilarity(String text1, String text2) {
        float[] vec1 = embed(text1);
        float[] vec2 = embed(text2);
        return cosineSimilarity(vec1, vec2);
    }

    /** Cosine similarity between two vectors */
    public static double cosineSimilarity(float[] vec1, float[] vec2) {
        double dot = 0.0, norm1 = 0.0, norm2 = 0.0;
        for (int i = 0; i < vec1.length; i++) {
            dot += vec1[i] * vec2[i];
            norm1 += vec1[i] * vec1[i];
            norm2 += vec2[i] * vec2[i];
        }
        if (norm1 == 0 || norm2 == 0) return 0.0;
        return dot / (Math.sqrt(norm1) * Math.sqrt(norm2));
    }

    /** Batch similarity matrix */
    public Map<String, Object> batchSimilarity(List<String> texts) {
        int n = texts.size();
        List<float[]> vectors = new ArrayList<>();
        for (String t : texts) vectors.add(embed(t));
        List<List<Double>> matrix = new ArrayList<>();
        for (int i = 0; i < n; i++) {
            List<Double> row = new ArrayList<>();
            for (int j = 0; j < n; j++) {
                row.add(i == j ? 1.0 : cosineSimilarity(vectors.get(i), vectors.get(j)));
            }
            matrix.add(row);
        }
        return Map.of(
            "texts", texts,
            "dimension", vectors.get(0).length,
            "similarityMatrix", matrix
        );
    }
}

3. EmbeddingController

package com.badao.ai.controller;

import com.badao.ai.service.EmbeddingService;
import org.springframework.ai.embedding.EmbeddingResponse;
import org.springframework.web.bind.annotation.*;
import java.util.*;

@RestController
public class EmbeddingController {
    private final EmbeddingService embeddingService;
    public EmbeddingController(EmbeddingService embeddingService) { this.embeddingService = embeddingService; }

    @GetMapping("/ai/embed")
    public Map<String, Object> embed(@RequestParam(value = "text", defaultValue = "Hello World") String text) {
        float[] vector = embeddingService.embed(text);
        return Map.of(
            "text", text,
            "dimension", vector.length,
            "vectorPreview", previewVector(vector, 10)
        );
    }

    @PostMapping("/ai/embed/similarity")
    public Map<String, Object> similarity(@RequestBody Map<String, String> req) {
        String t1 = req.getOrDefault("text1", "");
        String t2 = req.getOrDefault("text2", "");
        float[] v1 = embeddingService.embed(t1);
        float[] v2 = embeddingService.embed(t2);
        double sim = EmbeddingService.cosineSimilarity(v1, v2);
        return Map.of(
            "text1", t1,
            "text2", t2,
            "similarity", Math.round(sim * 10000.0) / 100.0,
            "dimension", v1.length
        );
    }

    @PostMapping("/ai/embed/batch-similarity")
    public Map<String, Object> batchSimilarity(@RequestBody Map<String, Object> req) {
        @SuppressWarnings("unchecked")
        List<String> texts = (List<String>) req.get("texts");
        if (texts == null || texts.isEmpty()) texts = List.of("苹果", "香蕉", "电脑", "笔记本");
        return embeddingService.batchSimilarity(texts);
    }

    @PostMapping("/ai/embed/batch")
    public Map<String, Object> batchEmbed(@RequestBody Map<String, Object> req) {
        @SuppressWarnings("unchecked")
        List<String> texts = (List<String>) req.get("texts");
        if (texts == null || texts.isEmpty()) texts = List.of("Hello", "World");
        EmbeddingResponse resp = embeddingService.embedForResponse(texts);
        List<Map<String, Object>> results = resp.getResults().stream()
            .map(r -> Map.of(
                "index", r.getIndex(),
                "dimension", r.getOutput().size(),
                "vectorPreview", previewVector(toFloatArray(r.getOutput()), 10)
            ))
            .toList();
        return Map.of(
            "texts", texts,
            "model", "ollama-nomic-embed-text",
            "results", results
        );
    }

    private static List<Float> previewVector(float[] vector, int max) {
        int limit = Math.min(vector.length, max);
        List<Float> preview = new ArrayList<>();
        for (int i = 0; i < limit; i++) preview.add(Math.round(vector[i] * 10000f) / 10000f);
        return preview;
    }

    private static float[] toFloatArray(List<Double> list) {
        float[] arr = new float[list.size()];
        for (int i = 0; i < list.size(); i++) arr[i] = list.get(i).floatValue();
        return arr;
    }
}

4. Front‑end test page (embedding-test.html)

<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8">
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  <title>Embedding Test</title>
  <style>/* minimal styling omitted for brevity */</style>
</head>
<body>
  <div class="container">
    <h1>🔢 Text Embedding</h1>
    <!-- Single text embedding UI -->
    <div class="card">
      <h2>📌 1. Single Text</h2>
      <textarea id="singleText" placeholder="Enter text...">人工智能正在改变世界</textarea>
      <button onclick="doEmbed()">🔍 Embed</button>
      <div class="result-box" id="singleResult"><span class="loading">Waiting…</span></div>
    </div>
    <!-- Pair similarity UI -->
    <div class="card">
      <h2>📊 2. Pair Similarity</h2>
      <textarea id="textA" placeholder="Text A">苹果是一种水果</textarea>
      <textarea id="textB" placeholder="Text B">香蕉也是一种水果</textarea>
      <button onclick="doSimilarity()">📐 Compute</button>
      <div class="result-box" id="similarityResult"><span class="loading">Waiting…</span></div>
    </div>
    <!-- Batch matrix UI -->
    <div class="card">
      <h2>🧮 3. Batch Matrix</h2>
      <textarea id="batchText" placeholder="One line per text">苹果
香蕉
电脑
手机</textarea>
      <button onclick="doBatchSimilarity()">🧮 Generate</button>
      <div class="result-box" id="batchResult"><span class="loading">Waiting…</span></div>
    </div>
  </div>
  <script>
    // helper functions omitted for brevity – they set example texts, call the REST APIs and render results
  </script>
</body>
</html>

5. Usage steps

Pull the embedding model : ollama pull nomic-embed-text Start the Spring Boot application : run SpringAiDemoApplication or the provided 一键启动.bat Open the test page : http://localhost:885/embedding-test.html Existing functionality (e.g., function-call-test.html) remains unchanged

6. API reference

GET /ai/embed?text=xxx

– single‑text vectorization POST /ai/embed/similarity – cosine similarity of two texts POST /ai/embed/batch-similarity – similarity matrix for a list of texts POST /ai/embed/batch – batch vectorization returning full responses

7. Maven pom (unchanged)

<project xmlns="http://maven.apache.org/POM/4.0.0">
  <modelVersion>4.0.0</modelVersion>
  <parent>
    <groupId>org.springframework.boot</groupId>
    <artifactId>spring-boot-starter-parent</artifactId>
    <version>3.2.5</version>
  </parent>
  <groupId>com.badao.ai</groupId>
  <artifactId>spring-ai-ollama-functioncall</artifactId>
  <version>1.0</version>
  <properties>
    <java.version>17</java.version>
    <spring-ai.version>1.0.0-M6</spring-ai.version>
    <spring-ai-alibaba.version>1.0.0-M6.1</spring-ai-alibaba.version>
  </properties>
  <dependencies>
    <dependency>
      <groupId>org.springframework.boot</groupId>
      <artifactId>spring-boot-starter-web</artifactId>
    </dependency>
    <dependency>
      <groupId>org.springframework.ai</groupId>
      <artifactId>spring-ai-ollama-spring-boot-starter</artifactId>
      <version>${spring-ai-alibaba.version}</version>
    </dependency>
  </dependencies>
  <repositories>
    <repository>
      <id>spring-milestones</id>
      <url>https://repo.spring.io/milestone</url>
    </repository>
  </repositories>
</project>

8. Test result