Redaction Techniques: How We Accomplish Critical Tasks
BeyondReview’s redaction capability leverages advanced visual classification technology to evaluate sensitive data and redact based on either word or pattern matching or on redacting certain zones within groupings of visually-similar documents. Called BeyondRedaction, this capability includes:
-
Automated grouping of documents to allow reviewers to learn the location of sensitive data.
-
Ability to target selection of sensitive data to be redacted based on either text values or page coordinates.
-
Defensible redaction of sensitive data.
This capability allow corporations to quickly and cost effectively redact sensitive data at a rate far exceeding traditional standard redaction technologies and processes.
BeyondRedaction: A Use Case Overview
12 bankruptcy courts redact over 412,000 documents in three months.
Summary
A report by Bankruptcy Court Special Master Karl Schieneman details how the BeyondRedactions technology enabled 12 bankruptcy courts to redact over 412,000 documents in under three months. The report notes that the productivity of the contract attorneys doing the redaction was almost 15 times normal redaction throughput, saving 19,000 hours of labor and hundreds of thousands of dollars of cost compared to manual approaches.

Background
The bankruptcy judge overseeing the 12 cases required Garlock Sealing Technologies, Inc., to pay for the cost of redacting social security numbers and other PII prior to production and appointed Karl Schieneman as the special master to oversee the redaction and production process. After reviewing available options, Schieneman elected to use technology provided by BeyondRecognition.
Technology
BeyondRedaction technology is able to group documents based on visual similarity. This process enables reviewers to learn where the PII, trade secret, privilege, etc. is contained in the different document types and greatly speeds review and makes redactions far more consistent.
As part of its process, BeyondRedaction catalogs where the graphical elements or glyphs are located on each page. BeyondRedaction’s global glyph catalog associates text values with various glyphs so that the system knows the page coordinates for the various text values. This cataloging enables BeyondRedaction to redact based on either the text values that were used, or based on page coordinates even if there were no text values associated with the glyphs.
Redaction Based on Text Coordinates

Redaction Based on Page Coordinates
