# Simulation Validation Report

Generated: 2026-05-07T14:00:00Z

## Dataset

- Records: 120
- Schema errors: 0
- Duplicate IDs: 0
- Average NLP confidence: 0.758
- Low-confidence records: 30
- High-priority records already routed to review: 36

## Priority Mix

{
  "high": 36,
  "low": 40,
  "medium": 44
}

## Review State Mix

{
  "needs_review": 54,
  "reviewed": 36,
  "new": 30
}

## Source Counts

{
  "grant_feed": 30,
  "paper_search": 26,
  "feedback_pulse": 35,
  "policy_monitor": 29
}

## Topic Counts

{
  "NLP": 21,
  "governance": 19,
  "learning analytics": 25,
  "funding": 30,
  "student feedback": 25
}

## Interpretation

The simulated dashboard is ready for a first case-study pass if schema errors and duplicate IDs are both zero.
The main review queue should focus on high-priority records and low-confidence records because those are the
items most likely to affect user decisions or require human judgment.