diff --git a/src/main/java/com/supervision/pdfqaserver/cache/PromptCache.java b/src/main/java/com/supervision/pdfqaserver/cache/PromptCache.java
index 2ac1f02..91938bd 100644
--- a/src/main/java/com/supervision/pdfqaserver/cache/PromptCache.java
+++ b/src/main/java/com/supervision/pdfqaserver/cache/PromptCache.java
@@ -19,6 +19,8 @@ public class PromptCache {
 
     public static final String CLASSIFY_TABLE = "CLASSIFY_TABLE";
 
+    public static final String EXTRACT_TABLE_TITLE = "EXTRACT_TABLE_TITLE";
+
     public static final Map<String, String> promptMap = new HashMap<>();
 
     static {
@@ -32,6 +34,7 @@ public class PromptCache {
         promptMap.put(TEXT_TO_CYPHER, TEXT_TO_CYPHER_PROMPT);
         promptMap.put(GENERATE_ANSWER, GENERATE_ANSWER_PROMPT);
         promptMap.put(CLASSIFY_TABLE, CLASSIFY_TABLE_PROMPT);
+        promptMap.put(EXTRACT_TABLE_TITLE, EXTRACT_TABLE_TITLE_PROMPT);
     }
 
 
@@ -152,7 +155,8 @@ public class PromptCache {
             1. 完全保留原始表头字段名称，不做任何中英文转换或修改
             2. 将每行数据转换为一个独立对象
             3. 所有数值保留原始格式（包括逗号分隔符和小数点）
-            4. 表格第一列作为主键字段
+            4. 去除数值中的转移符号
+            5. 表格第一列作为主键字段
                         
             **输出格式：**
             {
@@ -361,4 +365,18 @@ public class PromptCache {
             	请处理以下表格：
             	{}
             """;
+
+    private static final String EXTRACT_TABLE_TITLE_PROMPT = """
+            你是一个表格处理专家，直接给出结果，不用解释。
+            **任务**
+            - 从文本中挑选出表格的标题。
+            **说明**
+            - 文本内容是表格上部的一段文字。
+                        
+            **输出要求**：
+            - 直接给出结果，不要解释
+                        
+            **需要处理的文本**
+            {}
+            """;
 }
diff --git a/src/main/java/com/supervision/pdfqaserver/config/OllamaChatModelAspect.java b/src/main/java/com/supervision/pdfqaserver/config/OllamaChatModelAspect.java
index 9ebebdd..c2a7678 100644
--- a/src/main/java/com/supervision/pdfqaserver/config/OllamaChatModelAspect.java
+++ b/src/main/java/com/supervision/pdfqaserver/config/OllamaChatModelAspect.java
@@ -30,11 +30,11 @@ public class OllamaChatModelAspect {
         // 获取原始参数
         Object[] args = joinPoint.getArgs();
         // 如果是String类型的call方法，修改其参数
-        if (StrUtil.equals(signature, callStringMessage) && args.length > 0 && args[0] instanceof String originalPrompt) {
-            args[0] = originalPrompt + "/no_think";
+        if (StrUtil.equals(signature, callStringMessage) && args.length > 0) {
+            args[0] = args[0] + "\n /no_think";
         }
         // 执行原方法
-        Object result = joinPoint.proceed();
+        Object result = joinPoint.proceed(args);
         if (StrUtil.equals(model,"qwen3:30b-a3b") ) {
             if(StrUtil.equals(signature, callStringMessage)){
                 result = ((String) result).replaceAll("(?is)<think\\b[^>]*>(.*?)</think>", "").trim();
diff --git a/src/main/java/com/supervision/pdfqaserver/dto/ERAttributeDTO.java b/src/main/java/com/supervision/pdfqaserver/dto/ERAttributeDTO.java
index 267e676..e90f037 100644
--- a/src/main/java/com/supervision/pdfqaserver/dto/ERAttributeDTO.java
+++ b/src/main/java/com/supervision/pdfqaserver/dto/ERAttributeDTO.java
@@ -41,6 +41,15 @@ public class ERAttributeDTO {
     public ERAttributeDTO() {
     }
 
+    public ERAttributeDTO(TruncationErAttribute truncationErAttribute) {
+        this.id = truncationErAttribute.getId();
+        this.terId = truncationErAttribute.getTerId();
+        this.associationType = truncationErAttribute.getAssociationType();
+        this.attribute = truncationErAttribute.getAttribute();
+        this.value = truncationErAttribute.getValue();
+        this.dataType = truncationErAttribute.getDataType();
+    }
+
     public ERAttributeDTO(String attribute, String value, String dataType) {
         this.attribute = attribute;
         this.value = value;
diff --git a/src/main/java/com/supervision/pdfqaserver/dto/EREDTO.java b/src/main/java/com/supervision/pdfqaserver/dto/EREDTO.java
index 349b7c2..3707b85 100644
--- a/src/main/java/com/supervision/pdfqaserver/dto/EREDTO.java
+++ b/src/main/java/com/supervision/pdfqaserver/dto/EREDTO.java
@@ -2,6 +2,7 @@ package com.supervision.pdfqaserver.dto;
 
 import cn.hutool.core.collection.CollUtil;
 import cn.hutool.core.lang.UUID;
+import cn.hutool.core.util.RandomUtil;
 import cn.hutool.core.util.StrUtil;
 import com.alibaba.fastjson.JSONArray;
 import com.alibaba.fastjson.JSONObject;
@@ -36,6 +37,11 @@ public class EREDTO {
                 JSONObject nodeJson = (JSONObject) node;
                 String name = nodeJson.getString("name");
                 String type = nodeJson.getString("type");
+                if (StrUtil.hasBlank(name,type)){
+                    continue;
+                }
+                name = StrUtil.trim(name);
+                type = StrUtil.trim(type);
                 JSONObject attributes = nodeJson.getJSONObject("attributes");
                 List<ERAttributeDTO> erAttributeDTOS = new ArrayList<>();
                 if (CollUtil.isNotEmpty(attributes)){
@@ -60,22 +66,33 @@ public class EREDTO {
                 List<ERAttributeDTO> erAttributeDTOS = new ArrayList<>();
                 if (CollUtil.isNotEmpty(attributes)){
                     for (String key : attributes.keySet()) {
+                        if (StrUtil.isBlank(key)){
+                            continue;
+                        }
                         Object value = attributes.get(key);
+                        if (value instanceof String){
+                            if (StrUtil.isBlank((String) value)){
+                                continue;
+                            }
+                            value = StrUtil.trim((String) value);
+                        }
                         String valueString = attributes.getString(key);
                         ERAttributeDTO erAttributeDTO = new ERAttributeDTO(key, valueString, value instanceof Number?"1":"0");
                         erAttributeDTOS.add(erAttributeDTO);
                     }
                 }
-                if (StrUtil.isEmpty(source) || StrUtil.isEmpty(target)){
+                if (StrUtil.isBlank(source) || StrUtil.isBlank(target)){
                     log.warn("truncationId:{} relation:{} 关系中source or target is empty",truncationId,relationJson);
                     continue;
                 }
-                Optional<EntityExtractionDTO> sourceTypeOpt = entities.stream().filter(e -> StrUtil.equals(e.getName(), source)).findFirst();
+                final String sourceTrim = StrUtil.trim(source);
+                Optional<EntityExtractionDTO> sourceTypeOpt = entities.stream().filter(e -> StrUtil.equals(e.getName(), sourceTrim)).findFirst();
                 if (sourceTypeOpt.isEmpty()){
                     log.warn("truncationId:{} relation:{} 关系中source在实体中不存在",truncationId,relationJson);
                     continue;
                 }
-                Optional<EntityExtractionDTO> targetTypeOpt = entities.stream().filter(e -> StrUtil.equals(e.getName(), target)).findFirst();
+                final String targetTrim = StrUtil.trim(target);
+                Optional<EntityExtractionDTO> targetTypeOpt = entities.stream().filter(e -> StrUtil.equals(e.getName(), targetTrim)).findFirst();
                 if (targetTypeOpt.isEmpty()){
                     log.warn("truncationId:{} relation:{} 关系中target在实体中不存在",truncationId,relationJson);
                     continue;
@@ -108,12 +125,22 @@ public class EREDTO {
             EntityExtractionDTO entityExtractionDTO = new EntityExtractionDTO();
             entityExtractionDTO.setEntity("行");
             // 避免表格行名重复
-            entityExtractionDTO.setName("行-" + UUID.randomUUID());
+            entityExtractionDTO.setName("行-" +  RandomUtil.randomString(UUID.randomUUID().toString(), 10));
             entityExtractionDTO.setTruncationId(truncationId);
             List<ERAttributeDTO> erAttributeDTOS = new ArrayList<>();
             for (Map.Entry<String, Object> tableEntry : tableJson.entrySet()) {
                 String key = tableEntry.getKey();
+                if (StrUtil.isBlank(key)){
+                    continue;
+                }
+                key = StrUtil.trim(key);
                 Object value = tableEntry.getValue();
+                if (value instanceof String){
+                    if (StrUtil.isBlank(value.toString())){
+                        continue;
+                    }
+                    value = StrUtil.trim((String) value);
+                }
                 ERAttributeDTO erAttributeDTO = new ERAttributeDTO(key, value.toString(), value instanceof Number ? "1" : "0");
                 erAttributeDTOS.add(erAttributeDTO);
             }
diff --git a/src/main/java/com/supervision/pdfqaserver/dto/EntityExtractionDTO.java b/src/main/java/com/supervision/pdfqaserver/dto/EntityExtractionDTO.java
index e202b2c..ca97a98 100644
--- a/src/main/java/com/supervision/pdfqaserver/dto/EntityExtractionDTO.java
+++ b/src/main/java/com/supervision/pdfqaserver/dto/EntityExtractionDTO.java
@@ -1,5 +1,6 @@
 package com.supervision.pdfqaserver.dto;
 
+import com.supervision.pdfqaserver.domain.TruncationEntityExtraction;
 import lombok.Data;
 
 import java.util.ArrayList;
@@ -35,6 +36,13 @@ public class EntityExtractionDTO {
     public EntityExtractionDTO() {
     }
 
+    public EntityExtractionDTO(TruncationEntityExtraction entityExtraction) {
+        this.id = entityExtraction.getId();
+        this.truncationId = entityExtraction.getTruncationId();
+        this.entity = entityExtraction.getEntity();
+        this.name = entityExtraction.getName();
+    }
+
     public EntityExtractionDTO(String truncationId, String entity, String name, List<ERAttributeDTO> attributes) {
         this.truncationId = truncationId;
         this.entity = entity;
diff --git a/src/main/java/com/supervision/pdfqaserver/dto/RelationExtractionDTO.java b/src/main/java/com/supervision/pdfqaserver/dto/RelationExtractionDTO.java
index f01010d..212f51e 100644
--- a/src/main/java/com/supervision/pdfqaserver/dto/RelationExtractionDTO.java
+++ b/src/main/java/com/supervision/pdfqaserver/dto/RelationExtractionDTO.java
@@ -51,7 +51,17 @@ public class RelationExtractionDTO {
     public RelationExtractionDTO() {
     }
 
-    public RelationExtractionDTO(String truncationId,String source, String sourceType,String relation, String target,String targetType, List<ERAttributeDTO> attributes) {
+    public RelationExtractionDTO(TruncationRelationExtraction relationExtraction) {
+        this.id = relationExtraction.getId();
+        this.truncationId = relationExtraction.getTruncationId();
+        this.source = relationExtraction.getSource();
+        this.sourceType = relationExtraction.getSourceType();
+        this.relation = relationExtraction.getRelation();
+        this.target = relationExtraction.getTarget();
+        this.targetType = relationExtraction.getTargetType();
+    }
+
+    public RelationExtractionDTO(String truncationId, String source, String sourceType, String relation, String target, String targetType, List<ERAttributeDTO> attributes) {
         this.truncationId = truncationId;
         this.source = source;
         this.relation = relation;
diff --git a/src/main/java/com/supervision/pdfqaserver/service/DocumentTruncationService.java b/src/main/java/com/supervision/pdfqaserver/service/DocumentTruncationService.java
index 6a871a6..64f9de8 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/DocumentTruncationService.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/DocumentTruncationService.java
@@ -18,5 +18,11 @@ public interface DocumentTruncationService extends IService<DocumentTruncation>
 
     void deleteByDocumentId(String documentId);
 
+    void deleteByDocumentIds(List<String> documentIds);
+
     List<DocumentTruncation> queryByDocumentId(String documentId);
+
+    List<DocumentTruncation> queryByDocumentIds(List<String> documentIds);
+
+    List<DocumentTruncation> queryNotERETruncate(List<String> documentIds);
 }
diff --git a/src/main/java/com/supervision/pdfqaserver/service/KnowledgeGraphService.java b/src/main/java/com/supervision/pdfqaserver/service/KnowledgeGraphService.java
index 1a246d8..159f0e5 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/KnowledgeGraphService.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/KnowledgeGraphService.java
@@ -1,6 +1,8 @@
 package com.supervision.pdfqaserver.service;
 
 import com.supervision.pdfqaserver.dto.EREDTO;
+import com.supervision.pdfqaserver.dto.TruncateDTO;
+import java.util.List;
 
 /**
  * 知识图谱服务接口
@@ -14,12 +16,16 @@ public interface KnowledgeGraphService {
      */
     void generateGraph(String documentId);
 
+    void generateGraph(List<EREDTO> eredtoList);
+
+    List<EREDTO> truncateERE(List<TruncateDTO> truncateDTOS);
+
 
     /**
      * 重置图数据
-     * @param documentId 文档ID
+     * @param pdfId pdfId
      */
-    void resetGraphData(String documentId);
+    void resetGraphData(String pdfId);
 
     /**
      * 提交生成图任务
@@ -31,4 +37,6 @@ public interface KnowledgeGraphService {
 
 
     void saveERE(EREDTO eredto, String truncationId);
+
+    List<EREDTO> listPdfEREDTO(String pdfId);
 }
diff --git a/src/main/java/com/supervision/pdfqaserver/service/TruncationEntityExtractionService.java b/src/main/java/com/supervision/pdfqaserver/service/TruncationEntityExtractionService.java
index dd01403..7efe9f6 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/TruncationEntityExtractionService.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/TruncationEntityExtractionService.java
@@ -15,4 +15,6 @@ public interface TruncationEntityExtractionService extends IService<TruncationEn
     void saveERE(List<EntityExtractionDTO> entities);
 
     void deleteByTruncationId(String truncationId);
+
+    List<TruncationEntityExtraction> queryByTruncationIds(List<String> truncationIds);
 }
diff --git a/src/main/java/com/supervision/pdfqaserver/service/TruncationErAttributeService.java b/src/main/java/com/supervision/pdfqaserver/service/TruncationErAttributeService.java
index 5638327..0765f01 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/TruncationErAttributeService.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/TruncationErAttributeService.java
@@ -14,4 +14,6 @@ public interface TruncationErAttributeService extends IService<TruncationErAttri
     void  deleteByTerId(String terId);
 
     void deleteByTerIds(List<String> terIds);
+
+    List<TruncationErAttribute> queryByTerIds(List<String> terIds);
 }
diff --git a/src/main/java/com/supervision/pdfqaserver/service/TruncationRelationExtractionService.java b/src/main/java/com/supervision/pdfqaserver/service/TruncationRelationExtractionService.java
index b8ca062..d953431 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/TruncationRelationExtractionService.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/TruncationRelationExtractionService.java
@@ -16,4 +16,5 @@ public interface TruncationRelationExtractionService extends IService<Truncation
 
     void deleteByTruncationId(String truncationId);
 
+    List<TruncationRelationExtraction> queryByTruncationIds(List<String> documentIds);
 }
diff --git a/src/main/java/com/supervision/pdfqaserver/service/impl/ChinesEsToEnglishGeneratorImpl.java b/src/main/java/com/supervision/pdfqaserver/service/impl/ChinesEsToEnglishGeneratorImpl.java
index 66250b8..b867383 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/impl/ChinesEsToEnglishGeneratorImpl.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/impl/ChinesEsToEnglishGeneratorImpl.java
@@ -19,7 +19,8 @@ public class ChinesEsToEnglishGeneratorImpl implements ChinesEsToEnglishGenerato
     public String generate(String chinese) {
         log.info("generate:开始翻译: {}",chinese);
         String prompt = PromptCache.promptMap.get(CHINESE_TO_ENGLISH);
-        String response = ollamaChatModel.call(StrUtil.format(prompt, chinese));
+        String format = StrUtil.format(prompt, chinese);
+        String response = ollamaChatModel.call(format);
         log.info("generate:chinese:{}翻译结果: {}",chinese,response);
         return response;
     }
diff --git a/src/main/java/com/supervision/pdfqaserver/service/impl/DocumentTruncationServiceImpl.java b/src/main/java/com/supervision/pdfqaserver/service/impl/DocumentTruncationServiceImpl.java
index 05e89e4..4d8db68 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/impl/DocumentTruncationServiceImpl.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/impl/DocumentTruncationServiceImpl.java
@@ -10,6 +10,7 @@ import com.supervision.pdfqaserver.mapper.DocumentTruncationMapper;
 import org.springframework.stereotype.Service;
 import org.springframework.transaction.annotation.Transactional;
 
+import java.util.ArrayList;
 import java.util.List;
 
 /**
@@ -42,10 +43,31 @@ public class DocumentTruncationServiceImpl extends ServiceImpl<DocumentTruncatio
         this.lambdaUpdate().eq(DocumentTruncation::getDocumentId, documentId).remove();
     }
 
+    @Override
+    public void deleteByDocumentIds(List<String> documentIds) {
+        if (CollUtil.isEmpty(documentIds)){
+            return;
+        }
+        this.lambdaUpdate().in(DocumentTruncation::getDocumentId, documentIds).remove();
+    }
+
     @Override
     public List<DocumentTruncation> queryByDocumentId(String documentId) {
         return this.lambdaQuery().eq(DocumentTruncation::getDocumentId, documentId).list();
     }
+
+    @Override
+    public List<DocumentTruncation> queryByDocumentIds(List<String> documentIds) {
+        if (CollUtil.isEmpty(documentIds)){
+            return new ArrayList<>();
+        }
+        return this.lambdaQuery().in(DocumentTruncation::getDocumentId, documentIds).list();
+    }
+
+    @Override
+    public List<DocumentTruncation> queryNotERETruncate(List<String> documentIds) {
+        return null;
+    }
 }
 
 
diff --git a/src/main/java/com/supervision/pdfqaserver/service/impl/KnowledgeGraphServiceImpl.java b/src/main/java/com/supervision/pdfqaserver/service/impl/KnowledgeGraphServiceImpl.java
index f2e7dbb..79c1ccf 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/impl/KnowledgeGraphServiceImpl.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/impl/KnowledgeGraphServiceImpl.java
@@ -2,13 +2,12 @@ package com.supervision.pdfqaserver.service.impl;
 
 import cn.hutool.core.collection.CollUtil;
 import cn.hutool.core.date.TimeInterval;
+import cn.hutool.core.util.NumberUtil;
 import cn.hutool.core.util.StrUtil;
+import cn.hutool.json.JSONUtil;
 import com.supervision.pdfqaserver.constant.DomainMetaGenerationEnum;
-import com.supervision.pdfqaserver.domain.ChineseEnglishWords;
-import com.supervision.pdfqaserver.domain.DocumentTruncation;
-import com.supervision.pdfqaserver.domain.DomainMetadata;
+import com.supervision.pdfqaserver.domain.*;
 import com.supervision.pdfqaserver.dto.*;
-import com.supervision.pdfqaserver.domain.PdfAnalysisOutput;
 import com.supervision.pdfqaserver.service.*;
 import com.supervision.pdfqaserver.thread.KnowledgeGraphGenerateTreadPool;
 import lombok.RequiredArgsConstructor;
@@ -18,6 +17,7 @@ import org.springframework.stereotype.Service;
 import org.springframework.transaction.annotation.Transactional;
 import java.util.ArrayList;
 import java.util.List;
+import java.util.stream.Collectors;
 
 @Slf4j
 @Service
@@ -38,6 +38,10 @@ public class KnowledgeGraphServiceImpl implements KnowledgeGraphService {
 
     private final TruncationEntityExtractionService truncationEntityExtractionService;
 
+    private final TruncationRelationExtractionService truncationRelationExtractionService;
+
+    private final TruncationErAttributeService truncationErAttributeService;
+
     private final TruncationRelationExtractionService relationExtractionService;
 
     private final ChinesEsToEnglishGenerator chinesEsToEnglishGenerator;
@@ -65,24 +69,15 @@ public class KnowledgeGraphServiceImpl implements KnowledgeGraphService {
         // 对切分后的文档进行命名实体识别
         timer.start("doEre");
         log.info("开始命名实体识别...");
-        List<EREDTO> eredtoList = new ArrayList<>();
-        for (TruncateDTO truncateDTO : truncateDTOS) {
-            EREDTO eredto = null;
-            try {
-                eredto = tripleConversionPipeline.doEre(truncateDTO);
-            } catch (Exception e) {
-                log.error("命名实体识别失败,切分文档id:{}", truncateDTO.getId(), e);
-            }
-            if (null == eredto){
-                continue;
-            }
-            // 保存实体关系抽取结果
-            this.saveERE(eredto, truncateDTO.getId());
-            eredtoList.add(eredto);
-        }
+        List<EREDTO> eredtoList = truncateERE(truncateDTOS);
         log.info("命名实体识别完成,耗时:{}秒", timer.intervalSecond("doEre"));
 
-        // 合并实体关系抽取结果
+        generateGraph(eredtoList);
+
+    }
+
+    @Override
+    public void generateGraph(List<EREDTO> eredtoList) {
         log.info("开始合并实体关系抽取结果...");
         List<EREDTO> mergedList = tripleConversionPipeline.mergeEreResults(eredtoList);
         log.info("合并实体关系抽取结果完成,合并后个数:{}", mergedList.size());
@@ -137,23 +132,100 @@ public class KnowledgeGraphServiceImpl implements KnowledgeGraphService {
             if (CollUtil.isEmpty(eredto.getEntities()) && CollUtil.isEmpty(eredto.getRelations())){
                 continue;
             }
+            // 构造一个字典
+            allWords = getChineseEnglishWords(eredto);
+
             eredto.setEn(allWords);
-            tripleToCypherExecutor.saveERE(eredto);
+            try {
+                tripleToCypherExecutor.saveERE(eredto);
+            } catch (Exception e) {
+                log.info("生成cypher语句失败,切分文档id:{}", JSONUtil.toJsonStr(eredto), e);
+            }
         }
+    }
+
+    private static List<ChineseEnglishWords> getChineseEnglishWords(EREDTO eredto) {
+        List<ChineseEnglishWords> allWords;
+        allWords = eredto.getEntities().stream().flatMap(entity -> {
+            List<ChineseEnglishWords> collect = entity.getAttributes().stream().map(e -> {
+                ChineseEnglishWords words = new ChineseEnglishWords();
+                words.setChineseWord(e.getAttribute());
+                words.setEnglishWord(e.getAttribute());
+                return words;
+            }).collect(Collectors.toList());
+            ChineseEnglishWords words = new ChineseEnglishWords();
+            words.setChineseWord(entity.getEntity());
+            words.setEnglishWord(entity.getEntity());
+            collect.add(words);
+            return collect.stream();
+        }).collect(Collectors.toList());
+
+        eredto.getRelations().stream().flatMap(relation -> {
+            List<ChineseEnglishWords> words = relation.getAttributes().stream().map(e -> {
+                ChineseEnglishWords word = new ChineseEnglishWords();
+                word.setChineseWord(e.getAttribute());
+                word.setEnglishWord(e.getAttribute());
+                return word;
+            }).collect(Collectors.toList());
+            ChineseEnglishWords words1 = new ChineseEnglishWords();
+            words1.setChineseWord(relation.getRelation());
+            words1.setEnglishWord(relation.getRelation());
+            words.add(words1);
+            ChineseEnglishWords words2 = new ChineseEnglishWords();
+            words2.setChineseWord(relation.getSourceType());
+            words2.setEnglishWord(relation.getSourceType());
+            words.add(words2);
+            ChineseEnglishWords words3 = new ChineseEnglishWords();
+            words3.setChineseWord(relation.getTargetType());
+            words3.setEnglishWord(relation.getTargetType());
+            words.add(words3);
+            return words.stream();
+        }).forEach(allWords::add);
+
+        return allWords;
+    }
 
+    @Override
+    public List<EREDTO> truncateERE(List<TruncateDTO> truncateDTOS) {
+        List<EREDTO> eredtoList = new ArrayList<>();
+        int truncateSize = truncateDTOS.size();
+        int index = 1;
+        for (TruncateDTO truncateDTO : truncateDTOS) {
+            log.info("开始命名实体识别,切分文档id:{},识别进度:{}", truncateDTO.getId(), NumberUtil.formatPercent((index*1.0)/truncateSize, 2));
+            index++;
+            EREDTO eredto = null;
+            try {
+                eredto = tripleConversionPipeline.doEre(truncateDTO);
+            } catch (Exception e) {
+                log.error("命名实体识别失败,切分文档id:{}", truncateDTO.getId(), e);
+            }
+            if (null == eredto){
+                continue;
+            }
+            // 保存实体关系抽取结果
+            this.saveERE(eredto, truncateDTO.getId());
+            eredtoList.add(eredto);
+        }
+        return eredtoList;
     }
 
     @Override
     @Transactional(rollbackFor = Exception.class)
-    public void resetGraphData(String documentId) {
-        log.info("resetGraphData:重置知识图谱数据,documentId:{}", documentId);
-        List<DocumentTruncation> documentTruncations = documentTruncationService.queryByDocumentId(documentId);
+    public void resetGraphData(String pdfId) {
+        log.info("resetGraphData:重置知识图谱数据,pdfId:{}", pdfId);
+        List<PdfAnalysisOutput> pdfAnalysisOutputs = pdfAnalysisOutputService.queryByPdfId(Integer.valueOf(pdfId));
+        if (CollUtil.isEmpty(pdfAnalysisOutputs)){
+            log.info("没有找到pdfId为{}的pdf分析结果", pdfId);
+            return;
+        }
+        List<String> documentIds = pdfAnalysisOutputs.stream().map(p -> String.valueOf(p.getId())).toList();
+        List<DocumentTruncation> documentTruncations = documentTruncationService.queryByDocumentIds(documentIds);
         if (CollUtil.isEmpty(documentTruncations)){
-            log.info("没有找到文档切分数据,documentId:{},不用重置数据...", documentId);
+            log.info("没有找到文档切分数据,pdfId:{},不用重置数据...", pdfId);
             return;
         }
         // 删除切分数据
-        documentTruncationService.deleteByDocumentId(documentId);
+        documentTruncationService.deleteByDocumentIds(documentIds);
         for (DocumentTruncation documentTruncation : documentTruncations) {
             String truncationId = documentTruncation.getId();
             // 删除实体数据
@@ -161,7 +233,7 @@ public class KnowledgeGraphServiceImpl implements KnowledgeGraphService {
             // 删除关系数据
             relationExtractionService.deleteByTruncationId(truncationId);
         }
-        log.info("重置知识图谱数据完成,documentId:{}", documentId);
+        log.info("重置知识图谱数据完成,pdfId:{}", pdfId);
 
     }
 
@@ -214,4 +286,46 @@ public class KnowledgeGraphServiceImpl implements KnowledgeGraphService {
         relationExtractionService.saveERE(eredto.getRelations());
     }
 
+    @Override
+    public List<EREDTO> listPdfEREDTO(String pdfId) {
+
+        List<PdfAnalysisOutput> pdfAnalysisOutputs = pdfAnalysisOutputService.queryByPdfId(Integer.valueOf(pdfId));
+        if (CollUtil.isEmpty(pdfAnalysisOutputs)){
+            log.info("没有找到pdfId为{}的pdf分析结果", pdfId);
+            return new ArrayList<>();
+        }
+        List<String> documentIds = pdfAnalysisOutputs.stream().map(p -> p.getId().toString()).toList();
+        List<DocumentTruncation> documentTruncations = documentTruncationService.queryByDocumentIds(documentIds);
+        List<String> truncationIds = documentTruncations.stream().map(DocumentTruncation::getId).toList();
+        List<TruncationEntityExtraction> truncationEntityExtractions = truncationEntityExtractionService.queryByTruncationIds(truncationIds);
+
+        List<TruncationRelationExtraction> truncationRelationExtractions = truncationRelationExtractionService.queryByTruncationIds(truncationIds);
+
+        List<String> teIds = truncationEntityExtractions.stream().map(TruncationEntityExtraction::getId).toList();
+        List<String> trIds = truncationRelationExtractions.stream().map(TruncationRelationExtraction::getId).collect(Collectors.toList());
+        trIds.addAll(teIds);
+        List<TruncationErAttribute> truncationErAttributes = truncationErAttributeService.queryByTerIds(trIds);
+
+        List<EREDTO> eres = new ArrayList<>();
+        for (TruncationEntityExtraction entityExtraction : truncationEntityExtractions) {
+            EREDTO eredto = new EREDTO();
+            EntityExtractionDTO extractionDTO = new EntityExtractionDTO(entityExtraction);
+            List<ERAttributeDTO> attributes = truncationErAttributes.stream()
+                    .filter(t -> StrUtil.equals(entityExtraction.getId(), t.getTerId())).map(ERAttributeDTO::new).collect(Collectors.toList());
+            extractionDTO.setAttributes(attributes);
+            eredto.getEntities().add(extractionDTO);
+            eres.add(eredto);
+        }
+        for (TruncationRelationExtraction relationExtraction : truncationRelationExtractions) {
+            EREDTO eredto = new EREDTO();
+            RelationExtractionDTO extractionDTO = new RelationExtractionDTO(relationExtraction);
+            List<ERAttributeDTO> attributes = truncationErAttributes.stream()
+                    .filter(t -> StrUtil.equals(relationExtraction.getId(), t.getTerId())).map(ERAttributeDTO::new).collect(Collectors.toList());
+            extractionDTO.setAttributes(attributes);
+            eredto.getRelations().add(extractionDTO);
+            eres.add(eredto);
+        }
+        return eres;
+    }
+
 }
diff --git a/src/main/java/com/supervision/pdfqaserver/service/impl/TripleConversionPipelineImpl.java b/src/main/java/com/supervision/pdfqaserver/service/impl/TripleConversionPipelineImpl.java
index 263e0cc..a15ff2b 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/impl/TripleConversionPipelineImpl.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/impl/TripleConversionPipelineImpl.java
@@ -3,6 +3,7 @@ package com.supervision.pdfqaserver.service.impl;
 import cn.hutool.core.collection.CollUtil;
 import cn.hutool.core.lang.Assert;
 import cn.hutool.core.util.BooleanUtil;
+import cn.hutool.core.util.RandomUtil;
 import cn.hutool.core.util.StrUtil;
 import com.supervision.pdfqaserver.cache.PromptCache;
 import com.supervision.pdfqaserver.constant.LayoutTypeEnum;
@@ -62,9 +63,40 @@ public class TripleConversionPipelineImpl implements TripleConversionPipeline {
                     truncateDTOS.add(truncateDTO);
                 }
             } else if (LayoutTypeEnum.TABLE.getCode() == layoutType) {
-                // 如果是表格类型的布局，直接添加到列表中
-                TruncateDTO truncateDTO = new TruncateDTO(documentDTO);
-                truncateDTOS.add(truncateDTO);
+                // 如果是表格类型的布局，进行切分
+
+                // 提前抽取表名
+                TableTitleDTO tableTitleDTO = this.extractTableTitle(documentDTO.getTitle());
+                if (null != tableTitleDTO && StrUtil.isNotEmpty(tableTitleDTO.getTitle())){
+                    documentDTO.setTitle(tableTitleDTO.getTitle());
+                }else {
+                    // 生成一个默认的表
+                    documentDTO.setTitle("tableName-"+ RandomUtil.randomString(10));
+                }
+                List<String> tableRows = StrUtil.split(documentDTO.getContent(), "\n").stream().filter(StrUtil::isNotEmpty).collect(Collectors.toList());
+                if (tableRows.size()<5){
+                    TruncateDTO truncateDTO = new TruncateDTO(documentDTO);
+                    truncateDTOS.add(truncateDTO);
+                    continue;
+                }
+                String tableTitle = tableRows.get(0);
+                // 标题分割符
+                String tableTitleSplit = tableRows.get(1);
+                List<String> noTitleRows = tableRows.subList(2,tableRows.size()-1);
+                List<List<String>> rows = CollUtil.split(noTitleRows, 4);
+                for (List<String> row : rows) {
+                    StringBuilder sb = new StringBuilder();
+                    sb.append(tableTitle).append("\n");
+                    sb.append(tableTitleSplit).append("\n");
+                    for (String s : row) {
+                        sb.append(s).append("\n");
+                    }
+                    TruncateDTO truncateDTO = new TruncateDTO(documentDTO);
+                    truncateDTO.setContent(sb.toString());
+                    truncateDTOS.add(truncateDTO);
+                }
+
+
             } else {
                 log.info("sliceDocuments:错误的布局类型: {}", layoutType);
             }
@@ -89,9 +121,10 @@ public class TripleConversionPipelineImpl implements TripleConversionPipeline {
             if (classify){
                 return doTextEre(truncateDTO);
             }
+
             return doTableEre(truncateDTO);
         }
-        log.info("doEre:错误的布局类型: {}", truncateDTO.getLayoutType());
+        log.warn("doEre:错误的布局类型: {}", truncateDTO.getLayoutType());
         return null;
     }
 
@@ -118,7 +151,14 @@ public class TripleConversionPipelineImpl implements TripleConversionPipeline {
 
     @Override
     public TableTitleDTO extractTableTitle(String content) {
-        return null;
+        TableTitleDTO tableTitleDTO = new TableTitleDTO();
+        if (StrUtil.isEmpty(content)){
+            log.warn("extractTableTitle:内容为空");
+            return tableTitleDTO;
+        }
+        String table = PromptCache.promptMap.get(PromptCache.EXTRACT_TABLE_TITLE);
+        tableTitleDTO.setTitle(table);
+        return tableTitleDTO;
     }
 
     private EREDTO doTextEre(TruncateDTO truncateDTO) {
@@ -140,6 +180,7 @@ public class TripleConversionPipelineImpl implements TripleConversionPipeline {
         // 手动设置表格标题
         EntityExtractionDTO titleEntity = new EntityExtractionDTO();
         titleEntity.setEntity("表");
+        titleEntity.setTruncationId(truncateDTO.getId());
         titleEntity.setName(truncateDTO.getTitle());
         // 添加关系
         List<RelationExtractionDTO> relations = new ArrayList<>();
diff --git a/src/main/java/com/supervision/pdfqaserver/service/impl/TripleToCypherExecutorImpl.java b/src/main/java/com/supervision/pdfqaserver/service/impl/TripleToCypherExecutorImpl.java
index 0cb970f..d897037 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/impl/TripleToCypherExecutorImpl.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/impl/TripleToCypherExecutorImpl.java
@@ -62,6 +62,7 @@ public class TripleToCypherExecutorImpl implements TripleToCypherExecutor {
                 Map<String, Object> attributes = entity.getAttributes().stream().collect(Collectors.toMap(
                         ERAttributeDTO::getAttributeEn, ERAttributeDTO::getValue
                 ));
+                attributes.put("truncationId", entity.getTruncationId());
                 attributes.put("name", entity.getName());
                 log.info("保存节点{},属性:{}", entity.getEntityEn(),JSONUtil.toJsonStr(entity.getAttributes()));
                 List<Long> nodeIds = neo4jRepository.saveOrUpdateEntityNode(entity.getEntityEn(), "name", attributes);
@@ -86,6 +87,8 @@ public class TripleToCypherExecutorImpl implements TripleToCypherExecutor {
                 Map<String, Object> attributes = relation.getAttributes().stream().collect(Collectors.toMap(
                         ERAttributeDTO::getAttributeEn, ERAttributeDTO::getValue
                 ));
+                attributes.put("sourceType", relation.getSourceType());
+                attributes.put("truncationId", relation.getTruncationId());
                 for (Long sourceNodeId : sourceNodeIds) {
                     for (Long targetNodeId : targetNodeIds) {
                         if (sourceNodeId.equals(targetNodeId)) {
diff --git a/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationEntityExtractionServiceImpl.java b/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationEntityExtractionServiceImpl.java
index 8295f13..7671764 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationEntityExtractionServiceImpl.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationEntityExtractionServiceImpl.java
@@ -13,7 +13,7 @@ import com.supervision.pdfqaserver.service.TruncationErAttributeService;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
 import org.springframework.stereotype.Service;
-
+import java.util.ArrayList;
 import java.util.List;
 
 /**
@@ -65,6 +65,14 @@ public class TruncationEntityExtractionServiceImpl extends ServiceImpl<Truncatio
             this.lambdaUpdate().eq(TruncationEntityExtraction::getTruncationId, truncationId).remove();
         }
     }
+
+    @Override
+    public List<TruncationEntityExtraction> queryByTruncationIds(List<String> truncationIds) {
+        if (CollUtil.isEmpty(truncationIds)){
+            return new ArrayList<>();
+        }
+        return this.lambdaQuery().in(TruncationEntityExtraction::getTruncationId, truncationIds).list();
+    }
 }
 
 
diff --git a/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationErAttributeServiceImpl.java b/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationErAttributeServiceImpl.java
index d5f2735..a3bf6ed 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationErAttributeServiceImpl.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationErAttributeServiceImpl.java
@@ -7,6 +7,7 @@ import com.supervision.pdfqaserver.domain.TruncationErAttribute;
 import com.supervision.pdfqaserver.service.TruncationErAttributeService;
 import com.supervision.pdfqaserver.mapper.TruncationErAttributeMapper;
 import org.springframework.stereotype.Service;
+import java.util.ArrayList;
 import java.util.List;
 
 /**
@@ -34,6 +35,14 @@ public class TruncationErAttributeServiceImpl extends ServiceImpl<TruncationErAt
         }
         this.lambdaUpdate().in(TruncationErAttribute::getTerId, terIds).remove();
     }
+
+    @Override
+    public List<TruncationErAttribute> queryByTerIds(List<String> terIds) {
+        if (CollUtil.isEmpty(terIds)){
+            return new ArrayList<>();
+        }
+        return this.lambdaQuery().in(TruncationErAttribute::getTerId, terIds).list();
+    }
 }
 
 
diff --git a/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationRelationExtractionServiceImpl.java b/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationRelationExtractionServiceImpl.java
index b9e9168..1421a61 100644
--- a/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationRelationExtractionServiceImpl.java
+++ b/src/main/java/com/supervision/pdfqaserver/service/impl/TruncationRelationExtractionServiceImpl.java
@@ -13,6 +13,7 @@ import com.supervision.pdfqaserver.mapper.TruncationRelationExtractionMapper;
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
 import org.springframework.stereotype.Service;
+import java.util.ArrayList;
 import java.util.List;
 
 /**
@@ -60,6 +61,14 @@ public class TruncationRelationExtractionServiceImpl extends ServiceImpl<Truncat
             this.lambdaUpdate().eq(TruncationRelationExtraction::getTruncationId, truncationId).remove();
         }
     }
+
+    @Override
+    public List<TruncationRelationExtraction> queryByTruncationIds(List<String> documentIds) {
+        if (CollUtil.isEmpty(documentIds)){
+            return new ArrayList<>();
+        }
+        return this.lambdaQuery().in(TruncationRelationExtraction::getTruncationId, documentIds).list();
+    }
 }
 
 
diff --git a/src/main/resources/logback.xml b/src/main/resources/logback.xml
index 28b28c3..290c5e1 100644
--- a/src/main/resources/logback.xml
+++ b/src/main/resources/logback.xml
@@ -44,7 +44,7 @@
                 <maxFileSize>100MB</maxFileSize>
             </timeBasedFileNamingAndTriggeringPolicy>
             <!--日志文档保留天数-->
-            <maxHistory>2</maxHistory>
+            <maxHistory>30</maxHistory>
             <totalSizeCap>500MB</totalSizeCap>
         </rollingPolicy>
         <!-- 此日志文档只记录debug级别的 -->
@@ -72,7 +72,7 @@
                 <maxFileSize>100MB</maxFileSize>
             </timeBasedFileNamingAndTriggeringPolicy>
             <!--日志文档保留天数-->
-            <maxHistory>2</maxHistory>
+            <maxHistory>30</maxHistory>
             <totalSizeCap>1GB</totalSizeCap>
         </rollingPolicy>
         <!-- 此日志文档只记录info级别的 -->
@@ -99,7 +99,7 @@
                 <maxFileSize>100MB</maxFileSize>
             </timeBasedFileNamingAndTriggeringPolicy>
             <!--日志文档保留天数-->
-            <maxHistory>2</maxHistory>
+            <maxHistory>30</maxHistory>
             <totalSizeCap>500MB</totalSizeCap>
         </rollingPolicy>
         <!-- 此日志文档只记录warn级别的 -->
@@ -126,7 +126,7 @@
                 <maxFileSize>100MB</maxFileSize>
             </timeBasedFileNamingAndTriggeringPolicy>
             <!--日志文档保留天数-->
-            <maxHistory>2</maxHistory>
+            <maxHistory>30</maxHistory>
             <totalSizeCap>500MB</totalSizeCap>
         </rollingPolicy>
         <!-- 此日志文档只记录ERROR级别的 -->
diff --git a/src/test/java/com/supervision/pdfqaserver/PdfQaServerApplicationTests.java b/src/test/java/com/supervision/pdfqaserver/PdfQaServerApplicationTests.java
index 7766570..60c5253 100644
--- a/src/test/java/com/supervision/pdfqaserver/PdfQaServerApplicationTests.java
+++ b/src/test/java/com/supervision/pdfqaserver/PdfQaServerApplicationTests.java
@@ -1,5 +1,6 @@
 package com.supervision.pdfqaserver;
 
+import com.supervision.pdfqaserver.dto.EREDTO;
 import com.supervision.pdfqaserver.service.ChinesEsToEnglishGenerator;
 import com.supervision.pdfqaserver.service.KnowledgeGraphService;
 import com.supervision.pdfqaserver.service.TripleConversionPipeline;
@@ -22,7 +23,15 @@ class PdfQaServerApplicationTests {
     private KnowledgeGraphService knowledgeGraphService;
     @Test
     void generateGraphTest() {
-        knowledgeGraphService.generateGraph("1");
+        knowledgeGraphService.generateGraph("40");
+        log.info("finish...");
+    }
+
+    @Test
+    void testGenerateGraph2() {
+        List<EREDTO> eredtos = knowledgeGraphService.listPdfEREDTO("17");
+
+        knowledgeGraphService.generateGraph(eredtos);
         log.info("finish...");
     }