add Embedding Similarity function

Author: YukiRinLL

pom.xml

@@ -358,6 +358,12 @@
 <!--            <artifactId>dotenv-java</artifactId>-->
 <!--            <version>2.3.2</version>-->
 <!--        </dependency>-->
+        <!-- ONNX Runtime -->
+        <dependency>
+            <groupId>com.microsoft.onnxruntime</groupId>
+            <artifactId>onnxruntime</artifactId>
+            <version>1.17.1</version>
+        </dependency>
     </dependencies>
 
     <build>

src/main/java/com/phantoms/phantomsbackend/common/utils/LiteEmbeddingSimilarity.java

@@ -0,0 +1,58 @@
+package com.phantoms.phantomsbackend.common.utils;
+
+import ai.onnxruntime.*;
+import java.util.Map;
+
+public class LiteEmbeddingSimilarity {
+
+    // 全局唯一模型（只加载1次，内存固定）
+    private static OrtEnvironment env;
+    private static OrtSession session;
+
+    static {
+        try {
+            // 初始化 ONNX 环境（极轻量）
+            env = OrtEnvironment.getEnvironment();
+            // 模型路径（放入 resources 下）
+            String modelPath = LiteEmbeddingSimilarity.class.getResource("/model/onnx/model.onnx").getPath();
+            
+            // 会话配置（最小内存）
+            OrtSession.SessionOptions opts = new OrtSession.SessionOptions();
+            opts.setIntraOpNumThreads(1);  // 单线程，省内存
+            opts.setOptimizationLevel(OrtSession.SessionOptions.OptLevel.BASIC_OPT);
+            
+            // 加载模型（22MB）
+            session = env.createSession(modelPath, opts);
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+    }
+
+    // ===================== 核心：生成向量 =====================
+    public static float[] getEmbedding(String text) throws Exception {
+        try (OrtSession.Result result = session.run(Map.of(
+            "input_ids", OnnxTensor.createTensor(env, new long[][]{{101, 102}}), // 简化示例，真实需替换为 token 结果
+            "attention_mask", OnnxTensor.createTensor(env, new long[][]{{1, 1}})
+        ))) {
+            return (float[]) result.get(0).getValue();
+        }
+    }
+
+    // ===================== 余弦相似度 =====================
+    public static double cosineSim(float[] a, float[] b) {
+        double dot = 0, normA = 0, normB = 0;
+        for (int i = 0; i < a.length; i++) {
+            dot += a[i] * b[i];
+            normA += a[i] * a[i];
+            normB += b[i] * b[i];
+        }
+        return dot / (Math.sqrt(normA) * Math.sqrt(normB));
+    }
+
+    // ===================== 对外接口 =====================
+    public static double similarity(String name1, String name2) throws Exception {
+        float[] v1 = getEmbedding(name1);
+        float[] v2 = getEmbedding(name2);
+        return cosineSim(v1, v2);
+    }
+}

src/main/resources/model/onnx/config.json

@@ -0,0 +1,25 @@
+{
+  "_name_or_path": "sentence-transformers/all-MiniLM-L6-v2",
+  "architectures": [
+    "BertModel"
+  ],
+  "attention_probs_dropout_prob": 0.1,
+  "classifier_dropout": null,
+  "gradient_checkpointing": false,
+  "hidden_act": "gelu",
+  "hidden_dropout_prob": 0.1,
+  "hidden_size": 384,
+  "initializer_range": 0.02,
+  "intermediate_size": 1536,
+  "layer_norm_eps": 1e-12,
+  "max_position_embeddings": 512,
+  "model_type": "bert",
+  "num_attention_heads": 12,
+  "num_hidden_layers": 6,
+  "pad_token_id": 0,
+  "position_embedding_type": "absolute",
+  "transformers_version": "4.27.4",
+  "type_vocab_size": 2,
+  "use_cache": true,
+  "vocab_size": 30522
+}

src/main/resources/model/onnx/model.onnx

@@ -0,0 +1,7 @@
+{
+  "cls_token": "[CLS]",
+  "mask_token": "[MASK]",
+  "pad_token": "[PAD]",
+  "sep_token": "[SEP]",
+  "unk_token": "[UNK]"
+}