Senior analyst Nick Nunekpeku puts the Redact functionality within RelativityOne through its paces and shares his thoughts in this latest blog from the Discovery & Disclosure team
Finally, I had a chance to try my hands at using RelativityOne Redact on a Data Subject Access Request (DSAR) project. RelativityOne Redact is a tool that enables a user to apply markups to image, PDF and Excel files. Redactions could be applied either manually or automatically. The Redact application supports the use of rules to specify words or terms for redaction. This allows terms in document review sets to be redacted in batches instead of redacting one term at a time, thereby saving a significant amount of time. The Redact application also supports the use of regular expressions (RegEx) for finding words for redaction, using pattern-matching.
Setting up a Redact Project is straightforward. While logged into RelativityOne, navigate to the Redaction Projects tab and click on Create new project. On the New project pop-up, select Image, Spreadsheet or PDF depending on the document type in the redaction set.
The settings that are available on the Markup Project creation page depend on the type of redaction project being created (Image, Excel or PDF), but generally the settings for all file types are Project Name, Saved Search, Markup Set and Rules.
Under the Rules section, Redaction or Highlight can be selected, in addition to the Markup Reason, Markup Scope and Markup Subtype. Markup Scope refers to the Word, Character or Line that needs to be redacted. If Redaction is selected, then the available Markup SubTypes are Black, Text, Cross, White or Inverse. If Highlighting is selected then the available Markup SubTypes are Yellow, Green, Blue, Orange, Pink or Purple. The Word/Phrase for redaction is also specified under the Rules section, in addition to regular expressions.
It is worth noting that when Redaction is selected for Excels, and Word is chosen as the Markup Scope, the only Markup SubType that can be used is Text. The Markup Text field allows a user to specify a redaction text (e.g., “REDACTED”). However, if the Markup Scope selected is Cell, Row, Column, or Sheet then the other Markup SubType options (Black and White) can be used. In addition to Redaction and Highlight, Excel Redaction projects have two more options – Inverse (Redaction) and Inverse Highlight.
Now, what do I love about this application? The speed. And the fact that a user could populate a simple CSV file with the various parameters (Word/Phrase, RegEx, Markup Type, Markup SubType, Markup Text and Markup Scope) for the redaction project. Once the new project is created and saved, the CSV file can be uploaded to the application and then run.
The document population in my little DSAR project was less than 6,000, and the application was able to apply over 180,000 markups to about 4,500 documents using a CSV file containing just over 2,000 rules (a combination of names, email addresses, telephone numbers and hyperlinks). The whole process took just about 2 hours to complete.
Another feature of RelativityOne Redact is that it supports native file redactions, so there was no need to image all the documents in the redaction set. In this project, 3 separate saved searches were created (PDFs, Excels, and all other file types). PDFs and Excel files had redactions applied to the native documents. The other file types were imaged, and the redactions were applied to the images.
What is “not so great” about the Redact application? To me, the most important is the lack of a way to revert only some of the redactions applied. As it stands, reverting redactions affects all the words that have been redacted in a particular redact project, so there is no way to pick and choose only redactions applied to specific words or terms to be reverted. For example, my DSAR project required the redaction of all third-party email addresses and phone numbers. My initial plan was to redact all email addresses in the data set, and then revert just redactions applied to the email address(es) of the data subject. This could not be done because the application does not provide a way to update and upload another CSV file to the same redaction project before reverting the redactions. For this reason, I had to find a way to capture all the third-party email addresses within the data set, and then feed just those to the redact project.
In conclusion, RelativityOne Redact is a great tool for performing automated redactions. A lot of time is saved by using the application as compared to performing manual redaction. A feature that would provide a way for the uploaded CSV files to be amended so that only specific redactions to words or terms are reverted would be great.