Learn TAR

Educational resources for understanding TAR concepts and techniques.

Getting Started

TAR (Technology-Assisted Review) is a document review technique that uses machine learning to identify relevant documents in large datasets. This site provides practical tools and reference materials to help you understand and implement TAR effectively.

Key Concepts

  • Prevalence: The share of documents in a population that are actually responsive.
  • Recall: The share of truly responsive documents that were found by the workflow.
  • Precision: The share of documents marked responsive that were actually responsive.
  • Elusion: A sample-based estimate of what responsive material may remain in the unreviewed set.

Recommended Workflow

  1. Start with prevalence estimation to understand your dataset
  2. Use control sets to validate your workflow over time
  3. Monitor elusion rates to assess stopping decisions
  4. Report recall and precision for defensibility