Clustering, a feature of Relativity Analytics, is an unsupervised machine learning process that organizes documents by conceptual similarity. It analyzes the text in each document to identify key concepts and then groups documents that share similar concepts. This process helps reviewers quickly understand the main themes within a large, unfamiliar document set without prior input, such as search terms or example documents. The resulting clusters can be used to prioritize review, identify irrelevant data, or perform quality control checks.

A. Sorting by size: This is a simple filtering or sorting operation based on a metadata field (file size) and does not involve conceptual analysis.

C. Encrypting data: Encryption is a security measure used to protect data from unauthorized access and is unrelated to document analysis or organization.

D. Deleting files: Deleting files is a data removal action. While the insights from clustering might inform which documents to delete, clustering itself does not perform deletion.

- Relativity - Analytics. (n.d.). Clustering. Relativity Documentation. Retrieved from the RelativityOne documentation portal. The documentation states, "Clustering is a tool that organizes a document universe into conceptually similar groups, which is useful for quickly assessing your case data."

- Relativity - Analytics. (n.d.). Analytics. Relativity Documentation. Retrieved from the RelativityOne documentation portal. The overview section describes how Analytics features, including clustering, use conceptual analysis to group documents.