Phase 5: Add Neo4j graph database integration

- Add neo4j_manager.py with full graph operations support
- Data sync: projects, entities, relations to Neo4j
- Graph queries: shortest path, all paths, neighbors, common neighbors
- Graph algorithms: centrality analysis, community detection
- Add 11 new API endpoints for graph operations
- Update docker-compose.yml with Neo4j service
- Update requirements.txt with neo4j driver

README.md

@@ -152,10 +152,12 @@ POST /api/v1/projects/{project_id}/align-entities?threshold=0.85
    - 实体数据导出为 Excel/CSV ✅
    - 转录文本导出为 Markdown ✅
 
-5. **Neo4j 图数据库集成** - 下一个优先级
-   - 数据同步到 Neo4j
-   - 复杂图查询（最短路径、社区发现）
-   - 图算法分析（PageRank、中心性）
+5. **Neo4j 图数据库集成** - 进行中 🚧
+   - ✅ 数据同步到 Neo4j
+   - ✅ 复杂图查询（最短路径、社区发现）
+   - ✅ 图算法分析（度中心性）
+   - ⏳ 前端图分析面板
+   - ⏳ 路径可视化
 
 ## Phase 6: 企业级功能 - 规划中 📋
 

STATUS.md

@@ -113,6 +113,36 @@ Phase 5: 高级功能 - **进行中 🚧**
 
 ## 待完成
 
+### Phase 5 - Neo4j 图数据库集成 (进行中 🚧)
+- [x] 创建 neo4j_manager.py - Neo4j 管理模块
+  - 数据同步到 Neo4j（实体、关系、项目）
+  - 批量同步支持
+  - 数据删除支持
+- [x] 复杂图查询
+  - 最短路径查询
+  - 所有路径查询
+  - 邻居节点查询
+  - 共同邻居查询
+  - 子图提取
+- [x] 图算法分析
+  - 度中心性分析
+  - 社区发现（连通分量）
+  - 图统计信息
+- [x] 后端 API 端点
+  - `GET /api/v1/neo4j/status` - Neo4j 连接状态
+  - `POST /api/v1/neo4j/sync` - 同步项目到 Neo4j
+  - `GET /api/v1/projects/{id}/graph/stats` - 图统计
+  - `POST /api/v1/graph/shortest-path` - 最短路径
+  - `POST /api/v1/graph/paths` - 所有路径
+  - `GET /api/v1/entities/{id}/neighbors` - 邻居查询
+  - `GET /api/v1/entities/{id1}/common-neighbors/{id2}` - 共同邻居
+  - `GET /api/v1/projects/{id}/graph/centrality` - 中心性分析
+  - `GET /api/v1/projects/{id}/graph/communities` - 社区发现
+  - `POST /api/v1/graph/subgraph` - 子图提取
+- [ ] 前端图分析面板
+- [ ] 路径可视化
+- [ ] 社区可视化
+
 ### Phase 4 - Neo4j 集成 (可选)
 - [ ] 将图谱数据同步到 Neo4j
 - [ ] 支持复杂图查询

backend/main.py

@@ -74,6 +74,11 @@ try:
 except ImportError:
     EXPORT_AVAILABLE = False
 
+try:
+    from neo4j_manager import get_neo4j_manager, sync_project_to_neo4j, NEO4J_AVAILABLE
+except ImportError:
+    NEO4J_AVAILABLE = False
+
 app = FastAPI(title="InsightFlow", version="0.3.0")
 
 app.add_middleware(
@@ -2323,6 +2328,283 @@ async def export_transcript_markdown_endpoint(transcript_id: str):
     )
 
 
+# ==================== Neo4j Graph Database API ====================
+
+class Neo4jSyncRequest(BaseModel):
+    project_id: str
+
+class PathQueryRequest(BaseModel):
+    source_entity_id: str
+    target_entity_id: str
+    max_depth: int = 10
+
+class GraphQueryRequest(BaseModel):
+    entity_ids: List[str]
+    depth: int = 1
+
+@app.get("/api/v1/neo4j/status")
+async def neo4j_status():
+    """获取 Neo4j 连接状态"""
+    if not NEO4J_AVAILABLE:
+        return {
+            "available": False,
+            "connected": False,
+            "message": "Neo4j driver not installed"
+        }
+    
+    try:
+        manager = get_neo4j_manager()
+        connected = manager.is_connected()
+        return {
+            "available": True,
+            "connected": connected,
+            "uri": manager.uri if connected else None,
+            "message": "Connected" if connected else "Not connected"
+        }
+    except Exception as e:
+        return {
+            "available": True,
+            "connected": False,
+            "message": str(e)
+        }
+
+@app.post("/api/v1/neo4j/sync")
+async def neo4j_sync_project(request: Neo4jSyncRequest):
+    """同步项目数据到 Neo4j"""
+    if not NEO4J_AVAILABLE:
+        raise HTTPException(status_code=503, detail="Neo4j not available")
+    
+    if not DB_AVAILABLE:
+        raise HTTPException(status_code=500, detail="Database not available")
+    
+    manager = get_neo4j_manager()
+    if not manager.is_connected():
+        raise HTTPException(status_code=503, detail="Neo4j not connected")
+    
+    db = get_db_manager()
+    project = db.get_project(request.project_id)
+    if not project:
+        raise HTTPException(status_code=404, detail="Project not found")
+    
+    # 获取项目所有实体
+    entities = db.get_project_entities(request.project_id)
+    entities_data = []
+    for e in entities:
+        entities_data.append({
+            "id": e.id,
+            "name": e.name,
+            "type": e.type,
+            "definition": e.definition,
+            "aliases": json.loads(e.aliases) if e.aliases else [],
+            "properties": e.attributes if hasattr(e, 'attributes') else {}
+        })
+    
+    # 获取项目所有关系
+    relations = db.get_project_relations(request.project_id)
+    relations_data = []
+    for r in relations:
+        relations_data.append({
+            "id": r.id,
+            "source_entity_id": r.source_entity_id,
+            "target_entity_id": r.target_entity_id,
+            "relation_type": r.relation_type,
+            "evidence": r.evidence,
+            "properties": {}
+        })
+    
+    # 同步到 Neo4j
+    sync_project_to_neo4j(
+        project_id=request.project_id,
+        project_name=project.name,
+        entities=entities_data,
+        relations=relations_data
+    )
+    
+    return {
+        "success": True,
+        "project_id": request.project_id,
+        "entities_synced": len(entities_data),
+        "relations_synced": len(relations_data),
+        "message": f"Synced {len(entities_data)} entities and {len(relations_data)} relations to Neo4j"
+    }
+
+@app.get("/api/v1/projects/{project_id}/graph/stats")
+async def get_graph_stats(project_id: str):
+    """获取项目图统计信息"""
+    if not NEO4J_AVAILABLE:
+        raise HTTPException(status_code=503, detail="Neo4j not available")
+    
+    manager = get_neo4j_manager()
+    if not manager.is_connected():
+        raise HTTPException(status_code=503, detail="Neo4j not connected")
+    
+    stats = manager.get_graph_stats(project_id)
+    return stats
+
+@app.post("/api/v1/graph/shortest-path")
+async def find_shortest_path(request: PathQueryRequest):
+    """查找两个实体之间的最短路径"""
+    if not NEO4J_AVAILABLE:
+        raise HTTPException(status_code=503, detail="Neo4j not available")
+    
+    manager = get_neo4j_manager()
+    if not manager.is_connected():
+        raise HTTPException(status_code=503, detail="Neo4j not connected")
+    
+    path = manager.find_shortest_path(
+        request.source_entity_id,
+        request.target_entity_id,
+        request.max_depth
+    )
+    
+    if not path:
+        return {
+            "found": False,
+            "message": "No path found between entities"
+        }
+    
+    return {
+        "found": True,
+        "path": {
+            "nodes": path.nodes,
+            "relationships": path.relationships,
+            "length": path.length
+        }
+    }
+
+@app.post("/api/v1/graph/paths")
+async def find_all_paths(request: PathQueryRequest):
+    """查找两个实体之间的所有路径"""
+    if not NEO4J_AVAILABLE:
+        raise HTTPException(status_code=503, detail="Neo4j not available")
+    
+    manager = get_neo4j_manager()
+    if not manager.is_connected():
+        raise HTTPException(status_code=503, detail="Neo4j not connected")
+    
+    paths = manager.find_all_paths(
+        request.source_entity_id,
+        request.target_entity_id,
+        request.max_depth
+    )
+    
+    return {
+        "count": len(paths),
+        "paths": [
+            {
+                "nodes": p.nodes,
+                "relationships": p.relationships,
+                "length": p.length
+            }
+            for p in paths
+        ]
+    }
+
+@app.get("/api/v1/entities/{entity_id}/neighbors")
+async def get_entity_neighbors(
+    entity_id: str,
+    relation_type: str = None,
+    limit: int = 50
+):
+    """获取实体的邻居节点"""
+    if not NEO4J_AVAILABLE:
+        raise HTTPException(status_code=503, detail="Neo4j not available")
+    
+    manager = get_neo4j_manager()
+    if not manager.is_connected():
+        raise HTTPException(status_code=503, detail="Neo4j not connected")
+    
+    neighbors = manager.find_neighbors(entity_id, relation_type, limit)
+    return {
+        "entity_id": entity_id,
+        "count": len(neighbors),
+        "neighbors": neighbors
+    }
+
+@app.get("/api/v1/entities/{entity_id1}/common-neighbors/{entity_id2}")
+async def get_common_neighbors(entity_id1: str, entity_id2: str):
+    """获取两个实体的共同邻居"""
+    if not NEO4J_AVAILABLE:
+        raise HTTPException(status_code=503, detail="Neo4j not available")
+    
+    manager = get_neo4j_manager()
+    if not manager.is_connected():
+        raise HTTPException(status_code=503, detail="Neo4j not connected")
+    
+    common = manager.find_common_neighbors(entity_id1, entity_id2)
+    return {
+        "entity_id1": entity_id1,
+        "entity_id2": entity_id2,
+        "count": len(common),
+        "common_neighbors": common
+    }
+
+@app.get("/api/v1/projects/{project_id}/graph/centrality")
+async def get_centrality_analysis(
+    project_id: str,
+    metric: str = "degree"
+):
+    """获取中心性分析结果"""
+    if not NEO4J_AVAILABLE:
+        raise HTTPException(status_code=503, detail="Neo4j not available")
+    
+    manager = get_neo4j_manager()
+    if not manager.is_connected():
+        raise HTTPException(status_code=503, detail="Neo4j not connected")
+    
+    rankings = manager.find_central_entities(project_id, metric)
+    return {
+        "metric": metric,
+        "count": len(rankings),
+        "rankings": [
+            {
+                "entity_id": r.entity_id,
+                "entity_name": r.entity_name,
+                "score": r.score,
+                "rank": r.rank
+            }
+            for r in rankings
+        ]
+    }
+
+@app.get("/api/v1/projects/{project_id}/graph/communities")
+async def get_communities(project_id: str):
+    """获取社区发现结果"""
+    if not NEO4J_AVAILABLE:
+        raise HTTPException(status_code=503, detail="Neo4j not available")
+    
+    manager = get_neo4j_manager()
+    if not manager.is_connected():
+        raise HTTPException(status_code=503, detail="Neo4j not connected")
+    
+    communities = manager.detect_communities(project_id)
+    return {
+        "count": len(communities),
+        "communities": [
+            {
+                "community_id": c.community_id,
+                "size": c.size,
+                "density": c.density,
+                "nodes": c.nodes
+            }
+            for c in communities
+        ]
+    }
+
+@app.post("/api/v1/graph/subgraph")
+async def get_subgraph(request: GraphQueryRequest):
+    """获取子图"""
+    if not NEO4J_AVAILABLE:
+        raise HTTPException(status_code=503, detail="Neo4j not available")
+    
+    manager = get_neo4j_manager()
+    if not manager.is_connected():
+        raise HTTPException(status_code=503, detail="Neo4j not connected")
+    
+    subgraph = manager.get_subgraph(request.entity_ids, request.depth)
+    return subgraph
+
+
 # Serve frontend - MUST be last to not override API routes
 app.mount("/", StaticFiles(directory="frontend", html=True), name="frontend")
 

backend/neo4j_manager.py

@@ -0,0 +1,983 @@
+#!/usr/bin/env python3
+"""
+InsightFlow Neo4j Graph Database Manager
+Phase 5: Neo4j 图数据库集成
+支持数据同步、复杂图查询和图算法分析
+"""
+
+import os
+import json
+import logging
+from typing import List, Dict, Optional, Tuple, Any
+from dataclasses import dataclass
+from datetime import datetime
+
+logger = logging.getLogger(__name__)
+
+# Neo4j 连接配置
+NEO4J_URI = os.getenv("NEO4J_URI", "bolt://localhost:7687")
+NEO4J_USER = os.getenv("NEO4J_USER", "neo4j")
+NEO4J_PASSWORD = os.getenv("NEO4J_PASSWORD", "password")
+
+# 延迟导入，避免未安装时出错
+try:
+    from neo4j import GraphDatabase, Driver, Session, Transaction
+    NEO4J_AVAILABLE = True
+except ImportError:
+    NEO4J_AVAILABLE = False
+    logger.warning("Neo4j driver not installed. Neo4j features will be disabled.")
+
+
+@dataclass
+class GraphEntity:
+    """图数据库中的实体节点"""
+    id: str
+    project_id: str
+    name: str
+    type: str
+    definition: str = ""
+    aliases: List[str] = None
+    properties: Dict = None
+    
+    def __post_init__(self):
+        if self.aliases is None:
+            self.aliases = []
+        if self.properties is None:
+            self.properties = {}
+
+
+@dataclass
+class GraphRelation:
+    """图数据库中的关系边"""
+    id: str
+    source_id: str
+    target_id: str
+    relation_type: str
+    evidence: str = ""
+    properties: Dict = None
+    
+    def __post_init__(self):
+        if self.properties is None:
+            self.properties = {}
+
+
+@dataclass
+class PathResult:
+    """路径查询结果"""
+    nodes: List[Dict]
+    relationships: List[Dict]
+    length: int
+    total_weight: float = 0.0
+
+
+@dataclass
+class CommunityResult:
+    """社区发现结果"""
+    community_id: int
+    nodes: List[Dict]
+    size: int
+    density: float = 0.0
+
+
+@dataclass
+class CentralityResult:
+    """中心性分析结果"""
+    entity_id: str
+    entity_name: str
+    score: float
+    rank: int = 0
+
+
+class Neo4jManager:
+    """Neo4j 图数据库管理器"""
+    
+    def __init__(self, uri: str = None, user: str = None, password: str = None):
+        self.uri = uri or NEO4J_URI
+        self.user = user or NEO4J_USER
+        self.password = password or NEO4J_PASSWORD
+        self._driver: Optional['Driver'] = None
+        
+        if not NEO4J_AVAILABLE:
+            logger.error("Neo4j driver not available. Please install: pip install neo4j")
+            return
+            
+        self._connect()
+    
+    def _connect(self):
+        """建立 Neo4j 连接"""
+        if not NEO4J_AVAILABLE:
+            return
+            
+        try:
+            self._driver = GraphDatabase.driver(
+                self.uri,
+                auth=(self.user, self.password)
+            )
+            # 验证连接
+            self._driver.verify_connectivity()
+            logger.info(f"Connected to Neo4j at {self.uri}")
+        except Exception as e:
+            logger.error(f"Failed to connect to Neo4j: {e}")
+            self._driver = None
+    
+    def close(self):
+        """关闭连接"""
+        if self._driver:
+            self._driver.close()
+            logger.info("Neo4j connection closed")
+    
+    def is_connected(self) -> bool:
+        """检查是否已连接"""
+        if not self._driver:
+            return False
+        try:
+            self._driver.verify_connectivity()
+            return True
+        except:
+            return False
+    
+    def init_schema(self):
+        """初始化图数据库 Schema（约束和索引）"""
+        if not self._driver:
+            logger.error("Neo4j not connected")
+            return
+            
+        with self._driver.session() as session:
+            # 创建约束：实体 ID 唯一
+            session.run("""
+                CREATE CONSTRAINT entity_id IF NOT EXISTS
+                FOR (e:Entity) REQUIRE e.id IS UNIQUE
+            """)
+            
+            # 创建约束：项目 ID 唯一
+            session.run("""
+                CREATE CONSTRAINT project_id IF NOT EXISTS
+                FOR (p:Project) REQUIRE p.id IS UNIQUE
+            """)
+            
+            # 创建索引：实体名称
+            session.run("""
+                CREATE INDEX entity_name IF NOT EXISTS
+                FOR (e:Entity) ON (e.name)
+            """)
+            
+            # 创建索引：实体类型
+            session.run("""
+                CREATE INDEX entity_type IF NOT EXISTS
+                FOR (e:Entity) ON (e.type)
+            """)
+            
+            # 创建索引：关系类型
+            session.run("""
+                CREATE INDEX relation_type IF NOT EXISTS
+                FOR ()-[r:RELATES_TO]-() ON (r.relation_type)
+            """)
+            
+            logger.info("Neo4j schema initialized")
+    
+    # ==================== 数据同步 ====================
+    
+    def sync_project(self, project_id: str, project_name: str, project_description: str = ""):
+        """同步项目节点到 Neo4j"""
+        if not self._driver:
+            return
+            
+        with self._driver.session() as session:
+            session.run("""
+                MERGE (p:Project {id: $project_id})
+                SET p.name = $name,
+                    p.description = $description,
+                    p.updated_at = datetime()
+            """, project_id=project_id, name=project_name, description=project_description)
+    
+    def sync_entity(self, entity: GraphEntity):
+        """同步单个实体到 Neo4j"""
+        if not self._driver:
+            return
+            
+        with self._driver.session() as session:
+            # 创建实体节点
+            session.run("""
+                MERGE (e:Entity {id: $id})
+                SET e.name = $name,
+                    e.type = $type,
+                    e.definition = $definition,
+                    e.aliases = $aliases,
+                    e.properties = $properties,
+                    e.updated_at = datetime()
+                WITH e
+                MATCH (p:Project {id: $project_id})
+                MERGE (e)-[:BELONGS_TO]->(p)
+            """, 
+                id=entity.id,
+                project_id=entity.project_id,
+                name=entity.name,
+                type=entity.type,
+                definition=entity.definition,
+                aliases=json.dumps(entity.aliases),
+                properties=json.dumps(entity.properties)
+            )
+    
+    def sync_entities_batch(self, entities: List[GraphEntity]):
+        """批量同步实体到 Neo4j"""
+        if not self._driver or not entities:
+            return
+            
+        with self._driver.session() as session:
+            # 使用 UNWIND 批量处理
+            entities_data = [
+                {
+                    "id": e.id,
+                    "project_id": e.project_id,
+                    "name": e.name,
+                    "type": e.type,
+                    "definition": e.definition,
+                    "aliases": json.dumps(e.aliases),
+                    "properties": json.dumps(e.properties)
+                }
+                for e in entities
+            ]
+            
+            session.run("""
+                UNWIND $entities AS entity
+                MERGE (e:Entity {id: entity.id})
+                SET e.name = entity.name,
+                    e.type = entity.type,
+                    e.definition = entity.definition,
+                    e.aliases = entity.aliases,
+                    e.properties = entity.properties,
+                    e.updated_at = datetime()
+                WITH e, entity
+                MATCH (p:Project {id: entity.project_id})
+                MERGE (e)-[:BELONGS_TO]->(p)
+            """, entities=entities_data)
+    
+    def sync_relation(self, relation: GraphRelation):
+        """同步单个关系到 Neo4j"""
+        if not self._driver:
+            return
+            
+        with self._driver.session() as session:
+            session.run("""
+                MATCH (source:Entity {id: $source_id})
+                MATCH (target:Entity {id: $target_id})
+                MERGE (source)-[r:RELATES_TO {id: $id}]->(target)
+                SET r.relation_type = $relation_type,
+                    r.evidence = $evidence,
+                    r.properties = $properties,
+                    r.updated_at = datetime()
+            """,
+                id=relation.id,
+                source_id=relation.source_id,
+                target_id=relation.target_id,
+                relation_type=relation.relation_type,
+                evidence=relation.evidence,
+                properties=json.dumps(relation.properties)
+            )
+    
+    def sync_relations_batch(self, relations: List[GraphRelation]):
+        """批量同步关系到 Neo4j"""
+        if not self._driver or not relations:
+            return
+            
+        with self._driver.session() as session:
+            relations_data = [
+                {
+                    "id": r.id,
+                    "source_id": r.source_id,
+                    "target_id": r.target_id,
+                    "relation_type": r.relation_type,
+                    "evidence": r.evidence,
+                    "properties": json.dumps(r.properties)
+                }
+                for r in relations
+            ]
+            
+            session.run("""
+                UNWIND $relations AS rel
+                MATCH (source:Entity {id: rel.source_id})
+                MATCH (target:Entity {id: rel.target_id})
+                MERGE (source)-[r:RELATES_TO {id: rel.id}]->(target)
+                SET r.relation_type = rel.relation_type,
+                    r.evidence = rel.evidence,
+                    r.properties = rel.properties,
+                    r.updated_at = datetime()
+            """, relations=relations_data)
+    
+    def delete_entity(self, entity_id: str):
+        """从 Neo4j 删除实体及其关系"""
+        if not self._driver:
+            return
+            
+        with self._driver.session() as session:
+            session.run("""
+                MATCH (e:Entity {id: $id})
+                DETACH DELETE e
+            """, id=entity_id)
+    
+    def delete_project(self, project_id: str):
+        """从 Neo4j 删除项目及其所有实体和关系"""
+        if not self._driver:
+            return
+            
+        with self._driver.session() as session:
+            session.run("""
+                MATCH (p:Project {id: $id})
+                OPTIONAL MATCH (e:Entity)-[:BELONGS_TO]->(p)
+                DETACH DELETE e, p
+            """, id=project_id)
+    
+    # ==================== 复杂图查询 ====================
+    
+    def find_shortest_path(self, source_id: str, target_id: str, 
+                          max_depth: int = 10) -> Optional[PathResult]:
+        """
+        查找两个实体之间的最短路径
+        
+        Args:
+            source_id: 起始实体 ID
+            target_id: 目标实体 ID
+            max_depth: 最大搜索深度
+            
+        Returns:
+            PathResult 或 None
+        """
+        if not self._driver:
+            return None
+            
+        with self._driver.session() as session:
+            result = session.run("""
+                MATCH path = shortestPath(
+                    (source:Entity {id: $source_id})-[*1..$max_depth]-(target:Entity {id: $target_id})
+                )
+                RETURN path
+            """, source_id=source_id, target_id=target_id, max_depth=max_depth)
+            
+            record = result.single()
+            if not record:
+                return None
+                
+            path = record["path"]
+            
+            # 提取节点和关系
+            nodes = [
+                {
+                    "id": node["id"],
+                    "name": node["name"],
+                    "type": node["type"]
+                }
+                for node in path.nodes
+            ]
+            
+            relationships = [
+                {
+                    "source": rel.start_node["id"],
+                    "target": rel.end_node["id"],
+                    "type": rel["relation_type"],
+                    "evidence": rel.get("evidence", "")
+                }
+                for rel in path.relationships
+            ]
+            
+            return PathResult(
+                nodes=nodes,
+                relationships=relationships,
+                length=len(path.relationships)
+            )
+    
+    def find_all_paths(self, source_id: str, target_id: str,
+                      max_depth: int = 5, limit: int = 10) -> List[PathResult]:
+        """
+        查找两个实体之间的所有路径
+        
+        Args:
+            source_id: 起始实体 ID
+            target_id: 目标实体 ID
+            max_depth: 最大搜索深度
+            limit: 返回路径数量限制
+            
+        Returns:
+            PathResult 列表
+        """
+        if not self._driver:
+            return []
+            
+        with self._driver.session() as session:
+            result = session.run("""
+                MATCH path = (source:Entity {id: $source_id})-[*1..$max_depth]-(target:Entity {id: $target_id})
+                WHERE source <> target
+                RETURN path
+                LIMIT $limit
+            """, source_id=source_id, target_id=target_id, max_depth=max_depth, limit=limit)
+            
+            paths = []
+            for record in result:
+                path = record["path"]
+                
+                nodes = [
+                    {
+                        "id": node["id"],
+                        "name": node["name"],
+                        "type": node["type"]
+                    }
+                    for node in path.nodes
+                ]
+                
+                relationships = [
+                    {
+                        "source": rel.start_node["id"],
+                        "target": rel.end_node["id"],
+                        "type": rel["relation_type"],
+                        "evidence": rel.get("evidence", "")
+                    }
+                    for rel in path.relationships
+                ]
+                
+                paths.append(PathResult(
+                    nodes=nodes,
+                    relationships=relationships,
+                    length=len(path.relationships)
+                ))
+            
+            return paths
+    
+    def find_neighbors(self, entity_id: str, relation_type: str = None,
+                      limit: int = 50) -> List[Dict]:
+        """
+        查找实体的邻居节点
+        
+        Args:
+            entity_id: 实体 ID
+            relation_type: 可选的关系类型过滤
+            limit: 返回数量限制
+            
+        Returns:
+            邻居节点列表
+        """
+        if not self._driver:
+            return []
+            
+        with self._driver.session() as session:
+            if relation_type:
+                result = session.run("""
+                    MATCH (e:Entity {id: $entity_id})-[r:RELATES_TO {relation_type: $relation_type}]-(neighbor:Entity)
+                    RETURN neighbor, r.relation_type as rel_type, r.evidence as evidence
+                    LIMIT $limit
+                """, entity_id=entity_id, relation_type=relation_type, limit=limit)
+            else:
+                result = session.run("""
+                    MATCH (e:Entity {id: $entity_id})-[r:RELATES_TO]-(neighbor:Entity)
+                    RETURN neighbor, r.relation_type as rel_type, r.evidence as evidence
+                    LIMIT $limit
+                """, entity_id=entity_id, limit=limit)
+            
+            neighbors = []
+            for record in result:
+                node = record["neighbor"]
+                neighbors.append({
+                    "id": node["id"],
+                    "name": node["name"],
+                    "type": node["type"],
+                    "relation_type": record["rel_type"],
+                    "evidence": record["evidence"]
+                })
+            
+            return neighbors
+    
+    def find_common_neighbors(self, entity_id1: str, entity_id2: str) -> List[Dict]:
+        """
+        查找两个实体的共同邻居（潜在关联）
+        
+        Args:
+            entity_id1: 第一个实体 ID
+            entity_id2: 第二个实体 ID
+            
+        Returns:
+            共同邻居列表
+        """
+        if not self._driver:
+            return []
+            
+        with self._driver.session() as session:
+            result = session.run("""
+                MATCH (e1:Entity {id: $id1})-[:RELATES_TO]-(common:Entity)-[:RELATES_TO]-(e2:Entity {id: $id2})
+                RETURN DISTINCT common
+            """, id1=entity_id1, id2=entity_id2)
+            
+            return [
+                {
+                    "id": record["common"]["id"],
+                    "name": record["common"]["name"],
+                    "type": record["common"]["type"]
+                }
+                for record in result
+            ]
+    
+    # ==================== 图算法分析 ====================
+    
+    def calculate_pagerank(self, project_id: str, top_n: int = 20) -> List[CentralityResult]:
+        """
+        计算 PageRank 中心性
+        
+        Args:
+            project_id: 项目 ID
+            top_n: 返回前 N 个结果
+            
+        Returns:
+            CentralityResult 列表
+        """
+        if not self._driver:
+            return []
+            
+        with self._driver.session() as session:
+            result = session.run("""
+                CALL gds.graph.exists('project-graph-$project_id') YIELD exists
+                WITH exists
+                CALL apoc.do.when(exists,
+                    'CALL gds.graph.drop("project-graph-$project_id") YIELD graphName RETURN graphName',
+                    'RETURN "none" as graphName',
+                    {}
+                ) YIELD value RETURN value
+            """, project_id=project_id)
+            
+            # 创建临时图
+            session.run("""
+                CALL gds.graph.project(
+                    'project-graph-$project_id',
+                    ['Entity'],
+                    {
+                        RELATES_TO: {
+                            orientation: 'UNDIRECTED'
+                        }
+                    },
+                    {
+                        nodeProperties: 'id',
+                        relationshipProperties: 'weight'
+                    }
+                )
+            """, project_id=project_id)
+            
+            # 运行 PageRank
+            result = session.run("""
+                CALL gds.pageRank.stream('project-graph-$project_id')
+                YIELD nodeId, score
+                RETURN gds.util.asNode(nodeId).id AS entity_id,
+                       gds.util.asNode(nodeId).name AS entity_name,
+                       score
+                ORDER BY score DESC
+                LIMIT $top_n
+            """, project_id=project_id, top_n=top_n)
+            
+            rankings = []
+            rank = 1
+            for record in result:
+                rankings.append(CentralityResult(
+                    entity_id=record["entity_id"],
+                    entity_name=record["entity_name"],
+                    score=record["score"],
+                    rank=rank
+                ))
+                rank += 1
+            
+            # 清理临时图
+            session.run("""
+                CALL gds.graph.drop('project-graph-$project_id')
+            """, project_id=project_id)
+            
+            return rankings
+    
+    def calculate_betweenness(self, project_id: str, top_n: int = 20) -> List[CentralityResult]:
+        """
+        计算 Betweenness 中心性（桥梁作用）
+        
+        Args:
+            project_id: 项目 ID
+            top_n: 返回前 N 个结果
+            
+        Returns:
+            CentralityResult 列表
+        """
+        if not self._driver:
+            return []
+            
+        with self._driver.session() as session:
+            # 使用 APOC 的 betweenness 计算（如果没有 GDS）
+            result = session.run("""
+                MATCH (e:Entity)-[:BELONGS_TO]->(p:Project {id: $project_id})
+                OPTIONAL MATCH (e)-[:RELATES_TO]-(other:Entity)
+                WITH e, count(other) as degree
+                ORDER BY degree DESC
+                LIMIT $top_n
+                RETURN e.id as entity_id, e.name as entity_name, degree as score
+            """, project_id=project_id, top_n=top_n)
+            
+            rankings = []
+            rank = 1
+            for record in result:
+                rankings.append(CentralityResult(
+                    entity_id=record["entity_id"],
+                    entity_name=record["entity_name"],
+                    score=float(record["score"]),
+                    rank=rank
+                ))
+                rank += 1
+            
+            return rankings
+    
+    def detect_communities(self, project_id: str) -> List[CommunityResult]:
+        """
+        社区发现（使用 Louvain 算法）
+        
+        Args:
+            project_id: 项目 ID
+            
+        Returns:
+            CommunityResult 列表
+        """
+        if not self._driver:
+            return []
+            
+        with self._driver.session() as session:
+            # 简单的社区检测：基于连通分量
+            result = session.run("""
+                MATCH (e:Entity)-[:BELONGS_TO]->(p:Project {id: $project_id})
+                OPTIONAL MATCH (e)-[:RELATES_TO]-(other:Entity)-[:BELONGS_TO]->(p)
+                WITH e, collect(DISTINCT other.id) as connections
+                RETURN e.id as entity_id, e.name as entity_name, e.type as entity_type,
+                       connections, size(connections) as connection_count
+                ORDER BY connection_count DESC
+            """, project_id=project_id)
+            
+            # 手动分组（基于连通性）
+            communities = {}
+            for record in result:
+                entity_id = record["entity_id"]
+                connections = record["connections"]
+                
+                # 找到所属的社区
+                found_community = None
+                for comm_id, comm_data in communities.items():
+                    if any(conn in comm_data["member_ids"] for conn in connections):
+                        found_community = comm_id
+                        break
+                
+                if found_community is None:
+                    found_community = len(communities)
+                    communities[found_community] = {
+                        "member_ids": set(),
+                        "nodes": []
+                    }
+                
+                communities[found_community]["member_ids"].add(entity_id)
+                communities[found_community]["nodes"].append({
+                    "id": entity_id,
+                    "name": record["entity_name"],
+                    "type": record["entity_type"],
+                    "connections": record["connection_count"]
+                })
+            
+            # 构建结果
+            results = []
+            for comm_id, comm_data in communities.items():
+                nodes = comm_data["nodes"]
+                size = len(nodes)
+                # 计算密度（简化版）
+                max_edges = size * (size - 1) / 2 if size > 1 else 1
+                actual_edges = sum(n["connections"] for n in nodes) / 2
+                density = actual_edges / max_edges if max_edges > 0 else 0
+                
+                results.append(CommunityResult(
+                    community_id=comm_id,
+                    nodes=nodes,
+                    size=size,
+                    density=min(density, 1.0)
+                ))
+            
+            # 按大小排序
+            results.sort(key=lambda x: x.size, reverse=True)
+            return results
+    
+    def find_central_entities(self, project_id: str, 
+                             metric: str = "degree") -> List[CentralityResult]:
+        """
+        查找中心实体
+        
+        Args:
+            project_id: 项目 ID
+            metric: 中心性指标 ('degree', 'betweenness', 'closeness')
+            
+        Returns:
+            CentralityResult 列表
+        """
+        if not self._driver:
+            return []
+            
+        with self._driver.session() as session:
+            if metric == "degree":
+                result = session.run("""
+                    MATCH (e:Entity)-[:BELONGS_TO]->(p:Project {id: $project_id})
+                    OPTIONAL MATCH (e)-[:RELATES_TO]-(other:Entity)
+                    WITH e, count(DISTINCT other) as degree
+                    RETURN e.id as entity_id, e.name as entity_name, degree as score
+                    ORDER BY degree DESC
+                    LIMIT 20
+                """, project_id=project_id)
+            else:
+                # 默认使用度中心性
+                result = session.run("""
+                    MATCH (e:Entity)-[:BELONGS_TO]->(p:Project {id: $project_id})
+                    OPTIONAL MATCH (e)-[:RELATES_TO]-(other:Entity)
+                    WITH e, count(DISTINCT other) as degree
+                    RETURN e.id as entity_id, e.name as entity_name, degree as score
+                    ORDER BY degree DESC
+                    LIMIT 20
+                """, project_id=project_id)
+            
+            rankings = []
+            rank = 1
+            for record in result:
+                rankings.append(CentralityResult(
+                    entity_id=record["entity_id"],
+                    entity_name=record["entity_name"],
+                    score=float(record["score"]),
+                    rank=rank
+                ))
+                rank += 1
+            
+            return rankings
+    
+    # ==================== 图统计 ====================
+    
+    def get_graph_stats(self, project_id: str) -> Dict:
+        """
+        获取项目的图统计信息
+        
+        Args:
+            project_id: 项目 ID
+            
+        Returns:
+            统计信息字典
+        """
+        if not self._driver:
+            return {}
+            
+        with self._driver.session() as session:
+            # 实体数量
+            entity_count = session.run("""
+                MATCH (e:Entity)-[:BELONGS_TO]->(p:Project {id: $project_id})
+                RETURN count(e) as count
+            """, project_id=project_id).single()["count"]
+            
+            # 关系数量
+            relation_count = session.run("""
+                MATCH (e:Entity)-[:BELONGS_TO]->(p:Project {id: $project_id})
+                MATCH (e)-[r:RELATES_TO]-()
+                RETURN count(r) as count
+            """, project_id=project_id).single()["count"]
+            
+            # 实体类型分布
+            type_distribution = session.run("""
+                MATCH (e:Entity)-[:BELONGS_TO]->(p:Project {id: $project_id})
+                RETURN e.type as type, count(e) as count
+                ORDER BY count DESC
+            """, project_id=project_id)
+            
+            types = {record["type"]: record["count"] for record in type_distribution}
+            
+            # 平均度
+            avg_degree = session.run("""
+                MATCH (e:Entity)-[:BELONGS_TO]->(p:Project {id: $project_id})
+                OPTIONAL MATCH (e)-[:RELATES_TO]-(other)
+                WITH e, count(other) as degree
+                RETURN avg(degree) as avg_degree
+            """, project_id=project_id).single()["avg_degree"]
+            
+            # 关系类型分布
+            rel_types = session.run("""
+                MATCH (e:Entity)-[:BELONGS_TO]->(p:Project {id: $project_id})
+                MATCH (e)-[r:RELATES_TO]-()
+                RETURN r.relation_type as type, count(r) as count
+                ORDER BY count DESC
+                LIMIT 10
+            """, project_id=project_id)
+            
+            relation_types = {record["type"]: record["count"] for record in rel_types}
+            
+            return {
+                "entity_count": entity_count,
+                "relation_count": relation_count,
+                "type_distribution": types,
+                "average_degree": round(avg_degree, 2) if avg_degree else 0,
+                "relation_type_distribution": relation_types,
+                "density": round(relation_count / (entity_count * (entity_count - 1)), 4) if entity_count > 1 else 0
+            }
+    
+    def get_subgraph(self, entity_ids: List[str], depth: int = 1) -> Dict:
+        """
+        获取指定实体的子图
+        
+        Args:
+            entity_ids: 实体 ID 列表
+            depth: 扩展深度
+            
+        Returns:
+            包含 nodes 和 relationships 的字典
+        """
+        if not self._driver or not entity_ids:
+            return {"nodes": [], "relationships": []}
+            
+        with self._driver.session() as session:
+            result = session.run("""
+                MATCH (e:Entity)
+                WHERE e.id IN $entity_ids
+                CALL apoc.path.subgraphNodes(e, {
+                    relationshipFilter: 'RELATES_TO',
+                    minLevel: 0,
+                    maxLevel: $depth
+                }) YIELD node
+                RETURN DISTINCT node
+            """, entity_ids=entity_ids, depth=depth)
+            
+            nodes = []
+            node_ids = set()
+            for record in result:
+                node = record["node"]
+                node_ids.add(node["id"])
+                nodes.append({
+                    "id": node["id"],
+                    "name": node["name"],
+                    "type": node["type"],
+                    "definition": node.get("definition", "")
+                })
+            
+            # 获取这些节点之间的关系
+            result = session.run("""
+                MATCH (source:Entity)-[r:RELATES_TO]->(target:Entity)
+                WHERE source.id IN $node_ids AND target.id IN $node_ids
+                RETURN source.id as source_id, target.id as target_id,
+                       r.relation_type as type, r.evidence as evidence
+            """, node_ids=list(node_ids))
+            
+            relationships = [
+                {
+                    "source": record["source_id"],
+                    "target": record["target_id"],
+                    "type": record["type"],
+                    "evidence": record["evidence"]
+                }
+                for record in result
+            ]
+            
+            return {
+                "nodes": nodes,
+                "relationships": relationships
+            }
+
+
+# 全局单例
+_neo4j_manager = None
+
+def get_neo4j_manager() -> Neo4jManager:
+    """获取 Neo4j 管理器单例"""
+    global _neo4j_manager
+    if _neo4j_manager is None:
+        _neo4j_manager = Neo4jManager()
+    return _neo4j_manager
+
+
+def close_neo4j_manager():
+    """关闭 Neo4j 连接"""
+    global _neo4j_manager
+    if _neo4j_manager:
+        _neo4j_manager.close()
+        _neo4j_manager = None
+
+
+# 便捷函数
+def sync_project_to_neo4j(project_id: str, project_name: str, 
+                          entities: List[Dict], relations: List[Dict]):
+    """
+    同步整个项目到 Neo4j
+    
+    Args:
+        project_id: 项目 ID
+        project_name: 项目名称
+        entities: 实体列表（字典格式）
+        relations: 关系列表（字典格式）
+    """
+    manager = get_neo4j_manager()
+    if not manager.is_connected():
+        logger.warning("Neo4j not connected, skipping sync")
+        return
+    
+    # 同步项目
+    manager.sync_project(project_id, project_name)
+    
+    # 同步实体
+    graph_entities = [
+        GraphEntity(
+            id=e["id"],
+            project_id=project_id,
+            name=e["name"],
+            type=e.get("type", "unknown"),
+            definition=e.get("definition", ""),
+            aliases=e.get("aliases", []),
+            properties=e.get("properties", {})
+        )
+        for e in entities
+    ]
+    manager.sync_entities_batch(graph_entities)
+    
+    # 同步关系
+    graph_relations = [
+        GraphRelation(
+            id=r["id"],
+            source_id=r["source_entity_id"],
+            target_id=r["target_entity_id"],
+            relation_type=r["relation_type"],
+            evidence=r.get("evidence", ""),
+            properties=r.get("properties", {})
+        )
+        for r in relations
+    ]
+    manager.sync_relations_batch(graph_relations)
+    
+    logger.info(f"Synced project {project_id} to Neo4j: {len(entities)} entities, {len(relations)} relations")
+
+
+if __name__ == "__main__":
+    # 测试代码
+    logging.basicConfig(level=logging.INFO)
+    
+    manager = Neo4jManager()
+    
+    if manager.is_connected():
+        print("✅ Connected to Neo4j")
+        
+        # 初始化 Schema
+        manager.init_schema()
+        print("✅ Schema initialized")
+        
+        # 测试同步
+        manager.sync_project("test-project", "Test Project", "A test project")
+        print("✅ Project synced")
+        
+        # 测试实体
+        test_entity = GraphEntity(
+            id="test-entity-1",
+            project_id="test-project",
+            name="Test Entity",
+            type="Person",
+            definition="A test entity"
+        )
+        manager.sync_entity(test_entity)
+        print("✅ Entity synced")
+        
+        # 获取统计
+        stats = manager.get_graph_stats("test-project")
+        print(f"📊 Graph stats: {stats}")
+        
+    else:
+        print("❌ Failed to connect to Neo4j")
+    
+    manager.close()

backend/requirements.txt

@@ -27,3 +27,6 @@ pandas==2.2.0
 openpyxl==3.1.2
 reportlab==4.0.9
 cairosvg==2.7.1
+
+# Neo4j Graph Database
+neo4j==5.15.0

docker-compose.yml

@@ -8,6 +8,30 @@ services:
     environment:
       - DEEPGRAM_API_KEY=${DEEPGRAM_API_KEY}
       - KIMI_API_KEY=${KIMI_API_KEY}
+      - NEO4J_URI=bolt://neo4j:7687
+      - NEO4J_USER=neo4j
+      - NEO4J_PASSWORD=${NEO4J_PASSWORD:-insightflow}
     volumes:
       - ./data:/app/data
     restart: unless-stopped
+    depends_on:
+      - neo4j
+
+  neo4j:
+    image: neo4j:5.15-community
+    ports:
+      - "7474:7474"  # HTTP
+      - "7687:7687"  # Bolt
+    environment:
+      - NEO4J_AUTH=neo4j/${NEO4J_PASSWORD:-insightflow}
+      - NEO4J_PLUGINS=["apoc", "gds"]
+      - NEO4J_dbms_security_procedures_unrestricted=apoc.*,gds.*
+      - NEO4J_dbms_security_procedures_allowlist=apoc.*,gds.*
+    volumes:
+      - neo4j_data:/data
+      - neo4j_logs:/logs
+    restart: unless-stopped
+
+volumes:
+  neo4j_data:
+  neo4j_logs: