Continuous active learning is a technology-assisted review workflow in which the model retrains constantly as reviewers code, continually surfacing the...

Document review is the examination of collected documents for responsiveness, privilege, and issues, and it is the single largest cost in electronic...

First-pass review is the initial examination of documents for responsiveness and privilege, typically the highest-volume review stage.

Issue coding assigns documents to case-specific categories during review, enabling later retrieval by subject rather than keyword.

Managed review is document review conducted by a vendor-supplied team under defined protocols and quality metrics.

Quality control sampling examines a subset of coded documents to measure reviewer accuracy and consistency.

Responsiveness is whether a document falls within the scope of a discovery request and must therefore be produced.

A review platform is the software environment where documents are hosted, searched, coded, and prepared for production.

Reviewer overturn rate measures how often second-level review changes a first-pass determination, indicating review accuracy.

Elusion testing samples documents excluded from review to estimate how much responsive material the process missed.

Search term negotiation is the process by which parties agree on the keywords used to identify potentially responsive documents.

Search term validation tests whether agreed terms actually retrieve responsive documents and how much they miss.

Second-level review re-examines documents flagged during first-pass for privilege, significance, or coding difficulty.

Technology-assisted review uses machine learning to prioritize or classify documents for responsiveness, reducing the volume requiring manual attorney review.