This system provides a comprehensive web-based visualization interface that supports the complete workflow from knowledge graph construction to high-quality QA generation. It features modern frontend design with integrated real-time charts, interactive graph visualization, and intelligent data management capabilities.
Template File: templates/navigation.html
Core Features:
- 🎯 Modern Card-Style Navigation - Intuitive functional module selection interface
- 📊 Quick Status Overview - System status and recent activity display
- 🚀 One-Click Access - Quick access to all functional modules
- 📱 Responsive Design - Compatible with desktop and mobile devices
Use Cases: System entry point, feature overview, and quick navigation
Template File: templates/single_qa.html
Core Features:
-
📝 Entity Input Module
- Smart input field supporting entity names (e.g., "diabetes", "artificial intelligence")
- Real-time input validation and suggestions
-
⚙️ Parameter Control Panel
- Maximum nodes setting (recommended 100-400)
- Iteration count control (recommended 5-15)
- Sampling strategy selection (5 algorithms: mixed, max_chain, etc.)
-
📈 Real-time Visualization Module
- Left-side D3.js force-directed graph showing real-time graph construction
- Dynamic rendering of nodes and connections
- Interactive graph operations (zoom, drag, node click)
-
📋 Detailed Log Stream
- Right-side real-time log display with WebSocket updates
- Construction step tracking and error monitoring
- Trace ID support for issue diagnosis
-
🎯 QA Results Display
- Generated question-answer pairs after construction completion
- Answer quality analysis and metadata information
Use Cases: Single entity deep analysis, algorithm testing, parameter tuning validation
Template File: templates/batch_generation.html
Core Features:
-
📂 Multiple Data Source Support
- ✏️ Manual Input - Bulk entity input via text box with line separation
- 📁 File Upload - Multiple format support: .txt, .csv, .json, .jsonl
-
⚙️ Batch Configuration Panel
- Generation quantity control (supports 1-1000 entities)
- Sampling parameter settings (nodes, iterations, algorithm selection)
- Quality level settings (fast/standard/high-quality modes)
- QPS limits and concurrency control
-
📊 Real-time Monitoring Dashboard
- Current processing entity display
- Overall progress bar and percentage
- Success rate statistics and failure records
- Estimated completion time (ETA)
-
💾 Results Management Module
- Real-time preview of generated results
- One-click JSONL format download
- Save to evaluation dataset functionality
- Failed task retry mechanism
Use Cases: Large-scale QA dataset generation, educational training data preparation, enterprise knowledge base construction
Template File: templates/data_evaluation.html
Core Features:
-
🔍 Intelligent Evaluation Engine
- 🧠 DeepSeek R1 Model - High-quality answer generation based on latest reasoning model
- ⚖️ Smart Comparison - Automatic comparison between standard and predicted answers
- 📊 Multi-dimensional Assessment - Accuracy, similarity, language quality, logical consistency
-
📈 Evaluation Dimension Analysis
- Answer accuracy statistics
- Semantic similarity calculation
- Reasoning chain completeness check
- Knowledge coverage analysis
-
📊 Visualization Analysis Module
- Chart.js interactive charts
- Multi-dimensional radar charts
- Time series trend analysis
- Error type distribution statistics
-
💼 Dataset Management
- Upload evaluation datasets
- Batch evaluation task management
- Evaluation result download and export
- Historical evaluation record queries
Use Cases: QA quality assessment, model performance analysis, dataset quality control
Template File: templates/comparison_evaluation.html
Core Features:
-
🔄 A/B Testing Framework
- Parallel comparison of multiple datasets
- Performance comparison across different models
- Algorithm effectiveness comparative analysis
-
📊 Comparison Dimension Settings
- Custom evaluation metrics
- Weight allocation mechanisms
- Statistical significance testing
-
📈 Visual Comparison Reports
- Side-by-side comparison charts
- Difference heatmaps
- Performance improvement/degradation analysis
-
💾 Comparison History Management
- Historical comparison record queries
- Comparison result export
- Report generation and sharing
Use Cases: Algorithm performance comparison, model upgrade validation, quality improvement analysis
Template File: templates/data_management.html
Core Features:
-
📁 Dataset Management Center
- View, edit, and delete generated datasets
- Multi-format support (JSONL, JSON, CSV)
- Dataset merge and split functionality
-
🔍 Smart Content Preview
- Online QA content preview with pagination
- Keyword search and filtering capabilities
- Real-time data quality checking
-
📊 Quality Analysis Tools
- Dataset statistics (question count, average length, etc.)
- Quality distribution analysis and visualization
- Anomaly detection and marking
-
🔧 Data Preprocessing Tools
- Language detection and classification
- Information leakage detection
- Entity replacement and anonymization
- Duplicate data removal
-
📥 Batch Operations
- Batch format conversion
- Multi-file merge processing
- Bulk download and export
Use Cases: Data cleaning, quality control, format conversion, dataset maintenance
Template File: templates/domain_tags.html
Core Features:
-
🎯 Smart Auto-Tagging
- AI automatic entity domain identification (medical, technology, culture, education, etc.)
- LLM-based semantic understanding annotation
- Batch annotation processing support
-
✏️ Manual Correction Functions
- Manual annotation and editing interface
- Batch tag modification tools
- Annotation history and rollback
-
📊 Statistical Analysis Module
- Domain distribution statistical charts
- Annotation quality analysis
- Domain coverage assessment
-
🏷️ Tag Management System
- Predefined domain tag library (sports, academic, politics, entertainment, literature, culture, economics, technology, history, medical, etc.)
- Custom tag creation
- Tag hierarchy management
Use Cases: Dataset classification, domain specialization, content organization, knowledge categorization
Template File: templates/runs_qa_generation.html
Core Features:
-
📂 Run Record Management
- Historical construction run record viewing
- Run status and metadata display
- Failed task reprocessing
-
🔄 Secondary QA Generation
- Generate new QA based on existing graph data
- Secondary application of different sampling strategies
- QPS-controlled batch processing
-
📊 Performance Monitoring
- Real-time generation progress tracking
- Success rate and quality statistics
- Resource usage monitoring
-
💾 Result Integration
- Multi-run result merging
- Incremental update support
- Version management and rollback
Use Cases: Graph reuse, incremental generation, historical data utilization
Template File: templates/final_datasets.html
Core Features:
-
📋 Dataset Overview
- Final production dataset management
- Version control and tag management
- Dataset metadata display
-
🔍 Advanced Filtering
- Multi-dimensional data filtering
- Custom query conditions
- Real-time search and filtering
-
📊 Quality Assurance
- Unique ID management and validation
- Duplicate data detection
- Data integrity checking
-
📤 Production-Ready Export
- Multi-format export support
- Bulk download functionality
- Deployment-ready data package generation
Use Cases: Production data management, official releases, quality review, version publishing
- WebSocket Support - All pages support real-time status updates
- Progress Tracking - Real-time progress display for long-running tasks
- Error Monitoring - Real-time error capture and user alerts
- Responsive Design - Beautiful Bootstrap-based interface
- Interactive Charts - Chart.js and D3.js visualization components
- Intuitive Operations - Drag-and-drop, click, batch selection interactions
- Auto-save - Automatic result saving to prevent data loss
- Resume Capability - Support for interruption recovery of large tasks
- Version Management - Version tracking for data changes
- Parameter Validation - Real-time input parameter validation
- Error Handling - Comprehensive error capture and notification mechanisms
- Log Tracking - Detailed operation logs and Trace support
- Main Navigation → Understand system capabilities
- Single QA Testing → Familiarize with basic operations
- Batch Generation → Process actual data
- Data Evaluation → Validate result quality
- Data Management → Organize and optimize data
- Test First: Use single QA testing to validate parameter settings
- Batch Processing: Process large datasets in batches
- Quality Monitoring: Regularly use evaluation features to check data quality
- Backup Important Data: Download and backup important data promptly
- Parameter Tuning: Adjust generation parameters based on specific needs
Entity Preparation → Single QA Testing → Batch Generation → Quality Evaluation → Final Export
Literature Entities → Parameter Tuning → Large-scale Generation → Comparative Analysis → Publication
Domain Terms → Custom Tagging → Batch Processing → Quality Control → Production Deployment
- Increase parallel workers (recommended ≤5)
- Higher QPS limits (note API restrictions)
- Reduce max nodes and sample size
- Optimize model selection in configuration
- Increase sample size (recommended 15-20)
- Increase max nodes (recommended 250-400)
- Use max_chain sampling algorithm
- Increase max iterations
- Single worker with low QPS
- Smaller sample sizes
- Efficient model selection
- Batch processing scheduling
- JSONL - Standard question-answer pairs
- JSON - Structured data with metadata
- CSV - Tabular format for analysis
- TXT - Plain text for simple use cases
- REST APIs - All page functions accessible via REST
- WebSocket Events - Real-time status updates
- Batch Processing - Asynchronous task management
- Authentication - Secure access control
- Performance Metrics - Generation speed, success rates
- Quality Metrics - Accuracy, consistency, coverage
- Usage Statistics - User activity, resource utilization
- Custom Reports - Configurable analysis dashboards
- Connection Problems: Check WebSocket connectivity
- Performance Issues: Adjust concurrency settings
- Quality Problems: Review parameter configurations
- Data Issues: Use validation and cleaning tools
- System Logs: Detailed logging with trace IDs
- Error Messages: Clear error descriptions and solutions
- Documentation: Comprehensive user guides
- Community: User forums and knowledge base
📞 Technical Support: Check system logs or contact technical team for issues ⭐ Continuous Updates: System features are continuously optimized, check update notes regularly 🌟 Open Source: Contributions welcome - submit issues and pull requests!