质检

hwzj-api/src/main/java/api/controller/HomeController.java

@@ -149,26 +149,32 @@ public class HomeController extends BaseController {
     @ApiOperation("导出Excel")
     @Log(title = "导出excel", businessType = BusinessType.EXPORT)
     @PostMapping("/excelSentiment")
-    public  void  excelSentiment()  {
-
-        LambdaQueryWrapper<Label> lambdaQueryWrapper=new LambdaQueryWrapper<>();
-        lambdaQueryWrapper.ge(Label::getLabelId,16);
-        var list=labelService.getList(lambdaQueryWrapper);
-        List<ExcelSentiment> sentiments=new ArrayList<>();
-        if (list!=null&&list.size()>0)
+    public  void  excelSentiment() throws JsonProcessingException {
+        var mobile=translateService.getList();
+       List<ExcelSentiment> sentiments=new ArrayList<>();
+        if (mobile!=null&&mobile.size()>0)
         {
-            for(Label item:list)
+            ObjectMapper mapper = new ObjectMapper();
+            for(Translate item:mobile)
             {
-                ExcelSentiment excelSentiment=new ExcelSentiment();
-                excelSentiment.setContent(item.getLabelName());
-                if (item.getPatientCount() == 0L) {
-                    excelSentiment.setSeat("消极");
-                } else if (item.getPatientCount() == 2L) {
-                    excelSentiment.setSeat("积极");
-                } else {
-                    excelSentiment.setSeat("中立");
+                if(StringHelper.isNotEmpty(item.getTranslate()))
+                {
+                    RepTranslate [] translate = mapper.readValue(item.getTranslate(), RepTranslate[].class);
+                    for (RepTranslate iv :translate)
+                    {
+                        if(iv.getSpeech().length()>10&&iv.getNumber().length()<=4)
+                        {
+                            if(Objects.equals(modelService.OpenNLPDetection(iv.getSpeech()), "消极"))
+                            {
+                                ExcelSentiment excelSentiment=new ExcelSentiment();
+                                excelSentiment.setContent(iv.getSpeech());
+                                excelSentiment.setSeat("消极");
+                                sentiments.add(excelSentiment);
+                            }
+                        }
+                    }
                 }
-                sentiments.add(excelSentiment);
+
             }
         }
         ExcelHelper<ExcelSentiment> excel = new ExcelHelper<>(ExcelSentiment.class);

hwzj-api/src/main/java/api/controller/call/IvrController.java

@@ -62,7 +62,6 @@ public class IvrController extends BaseController {
         if (ivrService.getCount(qw) > 0) {
             return Error("此名称已经存在");
         }
-
         //input.setCreateBy(CurrentUser().getUserName());
         input.setCreateTime(new Date());
         input.setDelFlag("0");

hwzj-api/src/main/resources/application.yml

@@ -3,8 +3,8 @@ spring:
     date-format: yyyy-MM-dd HH:mm:ss
     time-zone: GMT+8
   profiles:
-    active: production
-    #active: dev
+    #active: production
+    active: dev
   servlet:
     multipart:
       max-file-size: 20MB #单个文件最大为20M

hwzj-service/pom.xml

@@ -47,7 +47,7 @@
             <artifactId>opennlp-tools</artifactId>
             <version>1.9.3</version>
         </dependency>
-
+    
 
 
     </dependencies>

hwzj-service/src/main/java/api/service/quality/impl/QcResultServiceImpl.java

@@ -105,13 +105,13 @@ public class QcResultServiceImpl extends BaseServiceImpl<QcResultMapper, QcResul
        //qw.apply("IFNULL(user_name,'') != ''");
         qw.apply("IFNULL(file_path,'') != ''");
          qw.eq(QcResult::getIsQuality,0);
-         if (!Objects.equals(callId, ""))
-         {
-             qw.eq(QcResult::getCallid,callId);
-         }
+//         if (!Objects.equals(callId, ""))
+//         {
+//             qw.eq(QcResult::getCallid,callId);
+//         }
        // qw.eq(QcResult::getUserName,"8002");
        // qw.ge (QcResult::getCallid,580);
-        qw.orderByDesc(QcResult::getQcId);
+        qw.orderByAsc(QcResult::getQcId);
         PageInput pageInput=new PageInput();
         pageInput.setPageNum(1);
         pageInput.setPageSize(100);
@@ -145,7 +145,7 @@ public class QcResultServiceImpl extends BaseServiceImpl<QcResultMapper, QcResul
                 String FileUrl="";String URL="";
                 //获取录音路径
                 if (!StringHelper.isEmpty(call.getFilePath())) {
-                    URL="http://10.0.0.3:9000/"+call.getFilePath();
+                    URL="http://1.194.161.64:9000/"+call.getFilePath();
                     FileUrl= downloadFile(URL,"files/luYin/"+ DateHelper.getDate());
                 }
 

hwzj-service/src/main/java/api/service/quality/impl/QualityModelServiceImpl.java

@@ -1,13 +1,8 @@
 package api.service.quality.impl;
-
-import api.entity.database.call.Translate;
-import api.entity.database.patient.Label;
-import api.entity.input.quality.RepTranslate;
-import api.entity.view.quality.ExcelSentiment;
+import api.entity.view.quality.TextSample;
 import api.service.call.ITranslateService;
 import api.service.patient.ILabelService;
 import api.service.quality.IQualityModelService;
-
 import opennlp.tools.doccat.*;
 import opennlp.tools.tokenize.SimpleTokenizer;
 import opennlp.tools.util.CollectionObjectStream;
@@ -15,14 +10,30 @@ import opennlp.tools.util.ObjectStream;
 import opennlp.tools.util.model.ModelUtil;
 import org.apache.poi.ss.usermodel.*;
 import org.apache.poi.xssf.usermodel.XSSFWorkbook;
-import org.apache.sis.util.resources.Vocabulary;
 
+import org.deeplearning4j.datasets.datavec.RecordReaderDataSetIterator;
+import org.deeplearning4j.nn.conf.MultiLayerConfiguration;
+import org.deeplearning4j.nn.conf.NeuralNetConfiguration;
+import org.deeplearning4j.nn.conf.layers.EmbeddingLayer;
+import org.deeplearning4j.nn.conf.layers.LSTM;
+import org.deeplearning4j.nn.conf.layers.RnnOutputLayer;
+import org.deeplearning4j.nn.weights.WeightInit;
+import org.nd4j.evaluation.classification.Evaluation;
+import org.nd4j.linalg.activations.Activation;
+
+import org.nd4j.linalg.dataset.api.iterator.DataSetIterator;
+import org.nd4j.linalg.learning.config.Adam;
+import org.nd4j.linalg.lossfunctions.LossFunctions;
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.stereotype.Service;
 import org.springframework.transaction.annotation.Transactional;
 import java.io.*;
 import java.util.*;
 
+
+
+
+
 @Transactional
 @Service
 public class QualityModelServiceImpl implements IQualityModelService {
@@ -58,7 +69,7 @@ public class QualityModelServiceImpl implements IQualityModelService {
             );
             // 3. 训练模型
             DoccatModel model = DocumentCategorizerME.train(
-                    "en",
+                    "zh",
                     sampleStream,
                     ModelUtil.createDefaultTrainingParameters(),
                     factory
@@ -80,7 +91,14 @@ public class QualityModelServiceImpl implements IQualityModelService {
     @Override
     public  String OpenNLPDetection(String text)
     {
-        return  GetSentiment(text);
+        if(isNumeric(text))
+        {
+            return "中立";
+        }
+        else
+        {
+            return  GetSentiment(text);
+        }
     }
     private String GetSentiment(String text) {
         try {
@@ -103,6 +121,13 @@ public class QualityModelServiceImpl implements IQualityModelService {
             return e.getMessage();
         }
     }
+    //判断是否为纯数字
+    public boolean isNumeric(String str) {
+        if (str == null || str.isEmpty()) {
+            return false;
+        }
+        return str.matches("[0-9]+");  // 匹配1个或多个数字
+    }
 
     private static List<DocumentSample> loadExcelData(String filePath) {
         List<DocumentSample> samples = new ArrayList<>();
@@ -150,21 +175,6 @@ public class QualityModelServiceImpl implements IQualityModelService {
 
 
 
-    // 文本向量化方法（需与训练时完全一致）
-    private double[] vectorizeText1(String text) {
-        // 此处应实现与训练一致的向量化逻辑
-        // 示例：简单词频统计（需替换为实际处理逻辑）
-        double[] features = new double[100];
-        String[] words = text.toLowerCase().split("\\s+");
-        for (String word : words) {
-            int index = Math.abs(word.hashCode()) % 100;
-            features[index] += 1.0;
-        }
-        return features;
-    }
-
-
-