What feature should you use to find documents that are conceptually correlated without regard to word count and order?

The feature "Find similar documents" is specifically designed to identify documents that share conceptual correlations irrespective of word count and order. This capability is particularly valuable in the context of legal or investigatory work, where the nuances of language can mean that documents expressing similar ideas may use different terms, structures, or length.

This feature utilizes advanced algorithms that analyze the content and context of documents, allowing it to detect relationships between documents based on their meaning rather than their textual similarities. This is crucial when trying to gather all relevant information on a topic without being limited by the exact phrasing or terminology used.

The other options have different applications. Textual near duplicates focus on identifying versions of documents that are largely identical in text. Keyword expansion typically expands the search by including synonyms or related terms, but it doesn’t analyze conceptual correlations in depth. Relativity Compare is used for comparing documents side by side to highlight differences but isn't aimed at finding related documents based on conceptual content. Thus, "Find similar documents" is the optimal choice for uncovering related materials meaningfully.