Files
ALwrity/docs-site/docs/features/ai/grounding-ui.md
2025-11-17 17:38:23 +05:30

335 lines
13 KiB
Markdown
Raw Permalink Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# Grounding UI
ALwrity's Grounding UI feature provides AI-powered content verification and fact-checking capabilities, ensuring your content is accurate, reliable, and trustworthy. This advanced feature helps maintain content credibility by grounding AI-generated content in verified information sources.
## Visuals
<p align="center">
<img src="../../assests/assistive-2.png" alt="Inline fact checking with citations and claim statuses" width="60%">
</p>
## What is Grounding UI?
Grounding UI is an intelligent content verification system that connects AI-generated content with real-world data sources, ensuring accuracy and reliability. It provides visual indicators, source citations, and verification status to help you create trustworthy, fact-checked content.
## Implementation Overview (Concise)
- Backend service: `backend/services/hallucination_detector.py` (claim extraction → evidence search → verification)
- Models: `backend/models/hallucination_models.py`
- API router: `backend/api/hallucination_detector.py` (registered in `backend/app.py`)
- Frontend service: `frontend/src/services/hallucinationDetectorService.ts`
- UI: LinkedIn Writer selection menu + FactCheckResults modal
### API Endpoints (Summary)
- `POST /api/hallucination-detector/detect` main fact-checking
- `POST /api/hallucination-detector/extract-claims` claims only
- `POST /api/hallucination-detector/verify-claim` single claim
- `GET /api/hallucination-detector/health` health check
### Minimal Setup
- Backend env:
- `EXA_API_KEY=...` (evidence search)
- `OPENAI_API_KEY=...` (claim extraction + verification)
- Frontend env:
- `REACT_APP_API_URL=http://localhost:8000`
### Quick Usage (UI)
1) In LinkedIn Writer, select a passage (10+ chars)
2) Click “Check Facts” in the selection menu
3) Review claims, assessments (supported/refuted/insufficient), confidence, and sources in the results modal
### Quick Usage (API)
```bash
curl -X POST "$API_URL/api/hallucination-detector/detect" \
-H "Content-Type: application/json" \
-d '{"text":"The Eiffel Tower is in Paris and built in 1889.","include_sources":true,"max_claims":5}'
```
### Key Benefits
- **Content Verification**: Verify facts and claims in real-time
- **Source Attribution**: Provide proper source citations
- **Credibility Enhancement**: Build trust with accurate content
- **Risk Mitigation**: Reduce misinformation and false claims
- **Quality Assurance**: Ensure content meets high standards
## Core Features
### Real-Time Fact Checking
#### Information Verification
- **Fact Validation**: Verify factual claims against reliable sources
- **Data Accuracy**: Check statistics and numerical data
- **Source Reliability**: Assess source credibility and authority
- **Claim Verification**: Validate specific claims and statements
- **Trend Analysis**: Verify current trends and developments
#### Source Integration
- **Multiple Sources**: Cross-reference information from multiple sources
- **Source Diversity**: Include various types of sources
- **Authority Assessment**: Evaluate source authority and expertise
- **Recency Check**: Ensure information is current and up-to-date
- **Bias Detection**: Identify potential bias in sources
### Visual Grounding Indicators
#### Verification Status
- **Verified**: Green indicators for verified information
- **Unverified**: Yellow indicators for unverified claims
- **Disputed**: Red indicators for disputed information
- **Outdated**: Orange indicators for outdated information
- **Source Missing**: Gray indicators for missing sources
#### Source Citations
- **Inline Citations**: Source links within content
- **Reference Lists**: Comprehensive reference sections
- **Source Types**: Different icons for different source types
- **Credibility Scores**: Visual credibility indicators
- **Last Updated**: Timestamp of last verification
### Content Enhancement
#### Accuracy Improvement
- **Fact Correction**: Suggest corrections for inaccurate information
- **Source Addition**: Recommend additional sources
- **Clarification**: Suggest clarifications for ambiguous content
- **Update Recommendations**: Suggest updates for outdated information
- **Bias Reduction**: Identify and suggest bias reduction
#### Trust Building
- **Transparency**: Show verification process and sources
- **Credibility Indicators**: Display content credibility scores
- **Expert Validation**: Highlight expert-reviewed content
- **Peer Review**: Show peer review status
- **Quality Metrics**: Display content quality indicators
## Integration with Research
### Web Research Integration
#### Real-Time Research
- **Live Data**: Access current information from web sources
- **Trend Monitoring**: Track real-time trends and developments
- **News Integration**: Include latest news and updates
- **Market Data**: Access current market information
- **Social Media**: Monitor social media discussions
#### Source Verification
- **Domain Authority**: Check website authority and credibility
- **Content Freshness**: Verify content recency
- **Author Credibility**: Assess author expertise and credentials
- **Publication Standards**: Evaluate publication quality
- **Fact-Checking Organizations**: Cross-reference with fact-checkers
### Database Integration
#### Knowledge Base
- **Internal Database**: Access internal knowledge base
- **Historical Data**: Reference historical information
- **Expert Knowledge**: Include expert-curated information
- **Industry Standards**: Reference industry best practices
- **Regulatory Information**: Include regulatory and compliance data
#### External Databases
- **Academic Sources**: Access academic and research databases
- **Government Data**: Include government and official sources
- **Industry Reports**: Reference industry research and reports
- **Statistical Databases**: Access statistical and data sources
- **News Archives**: Search historical news and information
## User Interface Features
### Visual Indicators
#### Status Icons
- **Checkmark**: Verified and accurate information
- **Warning**: Unverified or potentially inaccurate
- **Exclamation**: Disputed or controversial information
- **Clock**: Outdated or time-sensitive information
- **Question**: Missing or unclear information
#### Color Coding
- **Green**: Verified and reliable information
- **Yellow**: Requires verification or attention
- **Red**: Disputed or inaccurate information
- **Blue**: Source information and citations
- **Gray**: Neutral or informational content
### Interactive Elements
#### Hover Information
- **Source Details**: Show source information on hover
- **Verification Status**: Display verification details
- **Last Updated**: Show last verification timestamp
- **Credibility Score**: Display source credibility rating
- **Related Sources**: Show related source information
#### Click Actions
- **Source Links**: Direct links to source materials
- **Verification Details**: Detailed verification information
- **Source Analysis**: In-depth source analysis
- **Fact-Checking Reports**: Access to fact-checking reports
- **Update Requests**: Request content updates
## Content Types and Applications
### Blog Content
#### Fact-Checking
- **Statistical Claims**: Verify statistics and data
- **Historical Facts**: Check historical information
- **Expert Quotes**: Verify expert statements
- **Research Findings**: Validate research claims
- **Trend Analysis**: Verify trend information
#### Source Attribution
- **Research Citations**: Proper research citations
- **Expert References**: Expert opinion attribution
- **Data Sources**: Statistical data attribution
- **News References**: News and media citations
- **Industry Sources**: Industry-specific references
### Social Media Content
#### Quick Verification
- **Fact Checking**: Rapid fact verification
- **Source Links**: Quick access to sources
- **Credibility Indicators**: Visual credibility markers
- **Update Alerts**: Notifications for outdated information
- **Bias Warnings**: Bias detection and warnings
#### Engagement Enhancement
- **Trust Building**: Build audience trust
- **Transparency**: Show verification process
- **Credibility**: Enhance content credibility
- **Authority**: Establish thought leadership
- **Reliability**: Demonstrate content reliability
### Professional Content
#### Business Communications
- **Market Data**: Verify market information
- **Financial Data**: Check financial statistics
- **Regulatory Information**: Verify compliance data
- **Industry Standards**: Reference industry standards
- **Best Practices**: Validate best practice claims
#### Research and Analysis
- **Data Validation**: Verify research data
- **Methodology Review**: Check research methods
- **Source Evaluation**: Assess source quality
- **Bias Assessment**: Identify potential bias
- **Quality Assurance**: Ensure research quality
## Best Practices
### Content Creation
#### Verification Process
1. **Enable Grounding**: Activate grounding features
2. **Review Indicators**: Check verification status
3. **Verify Claims**: Ensure all claims are verified
4. **Add Sources**: Include proper source citations
5. **Update Regularly**: Keep information current
#### Quality Standards
1. **Source Diversity**: Use multiple source types
2. **Authority Sources**: Prioritize authoritative sources
3. **Current Information**: Ensure information is current
4. **Bias Awareness**: Be aware of potential bias
5. **Transparency**: Show verification process
### Source Management
#### Source Selection
- **Authority**: Choose authoritative sources
- **Recency**: Prefer recent information
- **Relevance**: Ensure source relevance
- **Diversity**: Include diverse perspectives
- **Credibility**: Verify source credibility
#### Citation Standards
- **Proper Attribution**: Give proper credit
- **Link Accessibility**: Ensure links work
- **Source Description**: Describe source context
- **Date Information**: Include publication dates
- **Author Information**: Include author details
## Advanced Features
### Customization Options
#### Verification Settings
- **Source Preferences**: Set preferred source types
- **Credibility Thresholds**: Define credibility standards
- **Update Frequency**: Set verification update frequency
- **Bias Sensitivity**: Adjust bias detection sensitivity
- **Quality Standards**: Set quality requirements
#### Display Options
- **Indicator Style**: Customize visual indicators
- **Color Schemes**: Choose color coding schemes
- **Information Density**: Adjust information display
- **Hover Details**: Customize hover information
- **Click Actions**: Set click action preferences
### Integration Features
#### API Integration
- **External APIs**: Connect to external data sources
- **Custom Databases**: Integrate custom databases
- **Real-Time Data**: Access real-time information
- **Automated Updates**: Enable automatic updates
- **Custom Sources**: Add custom source types
#### Workflow Integration
- **Content Review**: Integrate with content review process
- **Quality Gates**: Set quality checkpoints
- **Approval Workflow**: Include in approval process
- **Publishing Pipeline**: Integrate with publishing workflow
- **Performance Tracking**: Track verification performance
## Troubleshooting
### Common Issues
#### Verification Problems
- **Source Unavailable**: Handle unavailable sources
- **Outdated Information**: Manage outdated content
- **Bias Detection**: Address bias concerns
- **Credibility Issues**: Resolve credibility problems
- **Update Delays**: Manage update delays
#### Technical Issues
- **API Connectivity**: Resolve API connection issues
- **Data Synchronization**: Fix data sync problems
- **Performance Issues**: Address performance concerns
- **Display Problems**: Fix visual indicator issues
- **Integration Errors**: Resolve integration problems
### Environment & Health
- “EXA_API_KEY not found” → add key to backend `.env`, restart server
- “OpenAI API key not found” → add `OPENAI_API_KEY`, verify credits
- Health check: `GET /api/hallucination-detector/health`
### Getting Help
#### Support Resources
- **Documentation**: Review feature documentation
- **Tutorials**: Watch grounding UI tutorials
- **Best Practices**: Follow best practice guides
- **Community**: Join user community discussions
- **Support**: Contact technical support
#### Optimization Tips
- **Settings Review**: Regularly review settings
- **Source Management**: Maintain source quality
- **Update Monitoring**: Monitor update status
- **Performance Tracking**: Track verification performance
- **Continuous Improvement**: Continuously improve process
---
*Ready to enhance your content credibility with grounding UI? [Start with our First Steps Guide](../../getting-started/first-steps.md) and [Explore Blog Writer Features](../blog-writer/overview.md) to begin creating verified, trustworthy content!*