Much has been written about how to use advanced tools to select and produce electronic discovery. However, for every producing party there is a requesting party that has to review and analyze the produced documents.

Evaluating_Produced_DocsUnequal Footing. While producing parties have detailed knowledge about what was searched, the limitations of any tools used, and the criteria used for selection, requesting parties lack this insight. They have to get a handle on what was in the current production and ideally compare productions from the same party in related matters or compare productions from comparable litigation.

The Equalizer: Visual Classification. Visual classification provides a way to get an overview of what was produced in any specific production AND to compare the types of documents produced across different productions from the same or other parties. Document type overviews permit receiving parties to:

  • Zero in on document types most likely to contain relevant content.
  • Allocate document types to individual reviewers who have special knowledge in certain areas.
  • Conduct document-type specific searches for highly precise searching.
  • Compare produced scanned documents to see if the original native files were produced.

This is how it works: visual classification clusters visually-similar documents based on their appearance, not on the text associated with them. This means that scanned documents are classified the same as native electronic files regardless of the amount or quality of text associated with the documents.

Production Overview = Beginning Awareness. By reviewing one or two documents per cluster, attorneys can quickly gain a good sense of the type of documents in the production. Clusters can be arranged by the number of documents they contain, so that in just a few hours attorneys can have detailed familiarity with the types of documents produced and begin the analysis from a position of awareness..

Production Profiles Enable Comparisons. When reviewing clusters, attorneys can assign document-type labels to them and then obtain a summary or profile of how many of which types of documents were produced. The document-type profiles provide a way to compare productions from different parties or from the same parties for similar litigation.

Learning from Earlier Cases. Parties who regularly litigate similar issues with the same parties could also use visual classification to classify documents that were significant in earlier cases, e.g., deposition or trial exhibits, and make it a priority to examine the document clusters or document types used to classify those significant documents from earlier cases.

Excluding the Irrelevant. Alternatively, if requesting parties find that certain types of commonly-produced documents have no evidentiary value they can tell producing parties to omit producing them. Winnowing unwanted documents up front is one way to keep the costs associated with ediscovery reasonable and proportionate.

For more information on how you can use visual classification to evaluate incoming document productions, contact

Topics: Visual Classification, Ediscovery

Comments are closed.