DYNAMIC CLUSTERING

How It Works

Dynamic Clustering builds upon CAAT’s ability to identify relationships among documents based on concepts and extends it into a powerful document organization tool.

Dynamic Clustering is a virtually automatic function: CAAT is pointed toward a collection of documents or files and instructed to cluster them. CAAT creates a hierarchical file tree – based on concepts – and organizes conceptually-related files by grouping them into those tree nodes or clusters, even creating a title to describe what each cluster contains.

Users can easily vary the depth of the clusters (i.e., how many branches) as well as how generalized or specific those clustered documents can be in terms of conceptual similarity.

How it is Used in eDiscovery

With today’s staggering ESI volumes, many companies look to cull-down these collections before they begin applying any processing or review capabilities. Unfortunately, there’s little intuition behind most culling tools, so valuable documents are often discarded along with the nonresponsive ones.

Dynamic Clustering lets eDiscovery specialists rapidly organize ESI and identify any such important documents before they are lost to the culling process. Dynamic Clustering puts logic back into a collection of documents that is often based on timeframes or custodians – hardly logical when it comes to review.

Dynamic Clustering also helps speed the review process, by automatically putting conceptually-related, relevant documents together in a familiar file-tree structure, so knowledge workers can maintain continuity as they work with these documents.