Back in 2014, we first added redaction capabilities to the Alfresco Enterprise Viewer (then OpenAnnotate). Since that time the needs and new use-cases for digital redaction (like GDPR or FoIA) have only grown, and the redaction capabilities in the Enterprise Viewer have continue expanding to meet those needs. This post will be an overview of all the tools now available in the Alfresco Enterprise Viewer for efficient document redaction.
How has Redaction Evolved?
When Redaction was first added to the Enterprise Viewer the most pressing requirements were around security. Questions like, “Should I save an unredacted copy, and who can see it?” or “Is the text completely removed?”. With our initial focus on complete redaction of text and images to address security questions, the actual method of manually drawing redactions was relatively simple. As more involved use-cases like GDPR and FoIA requests evolved, the usage cases for redaction have increased. To meet this new demand two new types of redaction were added, automated redaction and bulk redaction.
Automated Suggestive Redaction
Originally added in 2018, suggestive redaction automatically uses patterns to find potential PII (Personally Identifiable Information) in a document that follow a certain format, like a Social Security Number. Text that follows these patterns is flagged for redaction and suggested to the user, who can then more efficiently redact the document.
More recent developed also allow for the automatic suggestion of properties. Since there may be metadata fields on the document or folder that commonly need redaction like Patient Name for a Patient Records scenario, these values can easy be found and suggested for redaction as well. Since these property-based redactions have a higher confidence, they make a strong candidate for totally automated redaction.
Bulk Redaction
In additional to more efficiently finding what to redact, GDPR has required the redaction of larger documents, claims or even entire repositories. For the latter two cases we added redaction capabilities to our other products, the Alfresco Content Accelerator and OpenMigrate.
In addition to these updates, we’ve added bulk redaction capabilities directly into the Enterprise Viewer to help with more document by document redaction cases like FoIA requests. A text search can be used to bulk redact any PII that was not caught by the automatic suggestions, optionally leaving a reason or exemption.
Combining the approaches
Both bulk and suggestive redaction drive the efficiency of the overall redaction process in different ways. To get the benefits of both, we recently updated the Enterprise Viewer grouping of suggested redactions. Suggestions highlighting the same PII (ex. same Social Security Number, or Name) in different places throughout the document are grouped, so they can be efficiently redaction at the same time. The grouped redaction suggestions comes closer to fully automated redaction, combining the speed of bulk redaction, with the efficient finding of information by suggestive redaction. See how each bulk, suggestive, and grouped suggestive redaction work in the video below:
What’s Next ?
In this post we’ve seen how the Alfresco Enterprise Viewer’s redaction has evolved over time to handle bigger cases like GDPR and FoIA. Next on the roadmap for redaction is further improve our Suggestive Redaction using the machine learning knowledge from our efforts to improve document indexing with Capture2.0.