优化得分项

virtual-patient-common/src/main/java/com/supervision/domain/QaSimilarityQuestionAnswer.java

@@ -0,0 +1,22 @@
+package com.supervision.domain;
+
+import lombok.Data;
+
+@Data
+public class QaSimilarityQuestionAnswer {
+
+    /**
+     * 匹配到的问题
+     */
+    private String matchQuestion;
+
+    /**
+     * 匹配到的问题ID
+     */
+    private String matchQuestionCode;
+
+    /**
+     * cosine余弦得分,一般0.5以上匹配度就比较高了
+     */
+    private Double matchScore;
+}

virtual-patient-common/src/main/java/com/supervision/util/SimilarityUtil.java

@@ -0,0 +1,56 @@
+package com.supervision.util;
+
+import com.supervision.domain.QaSimilarityQuestionAnswer;
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.ai.document.Document;
+import org.springframework.ai.vectorstore.RedisVectorStore;
+import org.springframework.ai.vectorstore.SearchRequest;
+
+import java.util.ArrayList;
+import java.util.Comparator;
+import java.util.List;
+import java.util.Optional;
+
+@Slf4j
+public class SimilarityUtil {
+
+    private static final RedisVectorStore redisVectorStore = SpringBeanUtil.getBean(RedisVectorStore.class);
+
+    /**
+     * 相似度比较,只找最高的
+     *
+     * @param question 问题
+     * @return 最高的TOP
+     */
+    public static Optional<QaSimilarityQuestionAnswer> talkRedisVectorWithScoreByFirst(String question) {
+        List<QaSimilarityQuestionAnswer> qaSimilarityQuestionAnswers = talkRedisVectorWithScore(question);
+        return qaSimilarityQuestionAnswers.stream().findFirst();
+    }
+
+    /**
+     * 相似度比较,找出所有的
+     *
+     * @param question 问题
+     * @return 相似度最高的top10
+     */
+    public static List<QaSimilarityQuestionAnswer> talkRedisVectorWithScore(String question) {
+        log.info("开始调用talkQaSimilarity,问题:{}", question);
+        try {
+            // 走Redis向量库进行比较,找出最高的top10
+            List<Document> documents = redisVectorStore.similaritySearch(SearchRequest
+                    .query(question)
+                    .withTopK(10));
+            return documents.stream().map(document -> {
+                QaSimilarityQuestionAnswer qaSimilarityQuestionAnswer = new QaSimilarityQuestionAnswer();
+                qaSimilarityQuestionAnswer.setMatchQuestion(document.getContent());
+                qaSimilarityQuestionAnswer.setMatchQuestionCode(String.valueOf(document.getMetadata().get("standardQuestionId")));
+                qaSimilarityQuestionAnswer.setMatchScore(Double.parseDouble(String.valueOf(document.getMetadata().get("vector_score"))));
+                return qaSimilarityQuestionAnswer;
+                // 排序,降序,取最高的
+            }).sorted(Comparator.comparing(QaSimilarityQuestionAnswer::getMatchScore).reversed()).toList();
+        } catch (Exception e) {
+            log.error("调用talkQaSimilarity error ", e);
+            return new ArrayList<>();
+        }
+    }
+}

virtual-patient-web/src/main/java/com/supervision/pojo/qaSimilarity/QaSimilarityQuestionAnswer.java

@@ -1,13 +0,0 @@
-package com.supervision.pojo.qaSimilarity;
-
-import lombok.Data;
-
-@Data
-public class QaSimilarityQuestionAnswer {
-
-    private String matchQuestion;
-
-    private String matchQuestionCode;
-
-    private Double matchScore;
-}

virtual-patient-web/src/main/java/com/supervision/pojo/vo/TalkVideoTtsResultResVO.java

@@ -1,6 +1,7 @@
 package com.supervision.pojo.vo;
 
 
+import com.fasterxml.jackson.annotation.JsonProperty;
 import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.Data;
 
@@ -17,6 +18,7 @@ public class TalkVideoTtsResultResVO {
     @Schema(description = "后端返回给前端时使用,表示该是语音回复还是action动作,1语音回复,2体格检查 3辅助检查")
     private Integer type = 1;
 
+    @JsonProperty(index = 100)
     @Schema(description = "音频base64位编码")
     private String voiceBase64;
 

virtual-patient-web/src/main/java/com/supervision/service/impl/AskServiceImpl.java

@@ -2,27 +2,24 @@ package com.supervision.service.impl;
 
 import cn.hutool.core.util.ObjectUtil;
 import cn.hutool.core.util.StrUtil;
-import cn.hutool.json.JSONUtil;
+import com.supervision.domain.QaSimilarityQuestionAnswer;
 import com.supervision.exception.BusinessException;
 import com.supervision.model.Process;
 import com.supervision.model.*;
-import com.supervision.pojo.qaSimilarity.QaSimilarityQuestionAnswer;
 import com.supervision.pojo.vo.TalkVideoReqVO;
 import com.supervision.pojo.vo.TalkVideoTtsResultResVO;
 import com.supervision.service.*;
 import com.supervision.util.AsrUtil;
+import com.supervision.util.SimilarityUtil;
 import com.supervision.util.TtsUtil;
 import com.supervision.util.UserUtil;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
-import org.springframework.ai.document.Document;
 import org.springframework.ai.vectorstore.RedisVectorStore;
 import org.springframework.beans.factory.annotation.Value;
 import org.springframework.stereotype.Service;
 import org.springframework.web.multipart.MultipartFile;
 
-import java.util.Comparator;
-import java.util.List;
 import java.util.Optional;
 
 @Slf4j
@@ -79,28 +76,6 @@ public class AskServiceImpl implements AskService {
         record.insert();
     }
 
-    public QaSimilarityQuestionAnswer talkRedisVectorWithScore(String question) {
-        log.info("开始调用talkQaSimilarity,问题:{}", question);
-        try {
-            // 走Redis进行比较
-            List<Document> documents = redisVectorStore.similaritySearch(question);
-            Optional<QaSimilarityQuestionAnswer> first = documents.stream().map(document -> {
-                QaSimilarityQuestionAnswer qaSimilarityQuestionAnswer = new QaSimilarityQuestionAnswer();
-                qaSimilarityQuestionAnswer.setMatchQuestion(document.getContent());
-                qaSimilarityQuestionAnswer.setMatchQuestionCode(String.valueOf(document.getMetadata().get("standardQuestionId")));
-                qaSimilarityQuestionAnswer.setMatchScore(Double.parseDouble(String.valueOf(document.getMetadata().get("vector_score"))));
-                return qaSimilarityQuestionAnswer;
-            }).max(Comparator.comparing(QaSimilarityQuestionAnswer::getMatchScore));
-            // 排序,降序,取最高的
-            log.info("调用talkQaSimilarity结束,问题:{},返回结果:{}", question, JSONUtil.toJsonStr(first.orElse(null)));
-            return first.orElse(null);
-        } catch (Exception e) {
-            log.error("调用talkQaSimilarity error ", e);
-            return null;
-        }
-    }
-
-
     /**
      * 使用无声视频+语音转文字的形式来做
      *
@@ -112,14 +87,15 @@ public class AskServiceImpl implements AskService {
         // 根据processId找到对应的病人
         Process process = Optional.ofNullable(processService.getById(talkReqVO.getProcessId())).orElseThrow(() -> new BusinessException("未找到诊疗进程"));
         MedicalRec medicalRec = medicalRecService.getById(process.getMedicalRecId());
-        QaSimilarityQuestionAnswer qaSimilarityQuestionAnswer = talkRedisVectorWithScore(talkReqVO.getText());
+        Optional<QaSimilarityQuestionAnswer> qaSimilarityQuestionAnswerOptional = SimilarityUtil.talkRedisVectorWithScoreByFirst(talkReqVO.getText());
         TalkVideoTtsResultResVO talkVideoTtsResultResVO = new TalkVideoTtsResultResVO();
         // 如果匹配度没有匹配到任何数据,则走大模型
-        if (ObjectUtil.isEmpty(qaSimilarityQuestionAnswer)) {
+        if (qaSimilarityQuestionAnswerOptional.isEmpty()) {
             String talk = aiService.talk(talkReqVO.getText(), medicalRec.getMedicalRecordAi());
             talkVideoTtsResultResVO.setAnswerMessage(talk);
             saveAiRecord(process.getId(), talkReqVO.getText(), talkVideoTtsResultResVO.getAnswerMessage());
         } else {
+            QaSimilarityQuestionAnswer qaSimilarityQuestionAnswer = qaSimilarityQuestionAnswerOptional.get();
             // 如果阈值过低,也走大模型
             double thresholdValue = Double.parseDouble(threshold);
             if (qaSimilarityQuestionAnswer.getMatchScore() < thresholdValue) {