Auto Redaction

Location: Analysis >> Auto Redaction

General

The Auto Redaction feature allows users to (i) redact multiple documents simultaneously, (ii) redact known and unknown content using custom and/or preset rules, and (iii) remove all redactions associated with a specific rule with the click of a button.

Currently, this feature can do the following:

  • Redact multiple documents simultaneously

  • Add multiple rules and filters to a job

  • Use Plain Text or Regular Expressions to identify content for redaction

  • Filter on Auto Redaction jobs and rules

  • Provides single click deletion of all redactions made by a specific rule

  • Clone, Rename, and/or Delete entire Auto Redaction jobs

Commonly Used Terms

Auto Redaction - A Lexbe tool that identifies and redacts multiple documents simultaneously using the rules set forth in an Auto Redaction job.

Regular Expression - A sequence of characters that specifies a search pattern (hereinafter, "Regex"). For example, if the letter d represents any single digit, then the Regex \b\d{3}-\d{2}-\d{4}\b would identify and redact all instances where the pattern ddd-dd-dddd is found (i.e. a Social Security Number).

Additional Resources

regexr.com - A website where users can build and test their custom Regular Expressions before applying them to their data.

Important Auto Redaction Insights

Please consider the following when using the Auto Redaction tool:

  • Auto Redaction searches the OCR text to identify what to redact (i.e. the text found in the TEXT tab of the Doc Viewer).

  • Low quality images and handwritten documents often have poor OCR which will negatively impact the accuracy of Auto Redaction.

  • At this time, Auto Redaction does not include the redaction of metadata.

  • Auto Redaction results should undergo QC prior to production. As such, we suggest running Auto Redaction prior to final document review, so QC can be done as part of that review process.

  • Auto Redaction renders a full redaction of any text that matches a rule (e.g. the entire SSN is redacted). Partial redactions (e.g. all but the last four digits of an SSN are redacted) can be achieved through the Custom Regular Expression and Custom Plain-Text options.

Creating an Auto Redaction Job

Click on the Analysis tab and select Auto Redaction from the drop down menu. To create a new Auto Redaction job, click the + button located at the top left corner of the screen. This will open a new job card. The job status will be set to "Created", and below that will be the current number of rules associated with the job.

Cloning an Auto Redaction Job

The cloning feature allows the user to begin crafting a new Auto Redaction job by generating a duplicate copy of any Auto Redaction job that’s already available in the case. To create a new Auto Redaction job via the cloning feature, click on the vertical ellipses ("kebab") of the job you wish to clone, and select Clone from the menu. This will create a new Auto Redaction job with any filters and rules from cloned job automatically populated.

Renaming an Auto Redaction Job

Auto Redaction jobs can be renamed at any time by either clicking on the job title and editing it as desired, or by clicking the kebab in the upper right corner, and selecting "Rename" from the drop down menu.

Adding Filters

Users can run Auto Redaction across the entire case, or they can add filters to run it across a specific subset of data. To add filters to an Auto Redaction job, users can click "Add Filter" in the job card, or click the kebab and select "Filter" from the drop down menu. Either option will open the Filter menu where you can select and apply any filters you wish to include in the current Auto Redaction job. This is the same Filter menu available on the Browse and Search pages, and, as such, it includes all built-in and custom fields in the case. Click here to read more on Lexbe's Simple and Advanced Filters.

Adding, Modifying, and Deleting Auto Redaction Rules Prior to Processing

  1. Click on the + button in the upper right corner to open the Add or Modify Rule window. The Add or Modify Rule window contains the following fields:

Select Preset or Custom - Use this dropdown menu to select the type of rule. Available options are:

  • Custom Plain-Text: Redacts all text that matches the text provided by the user. Use this option to redact specific known content (e.g. a known email address, SSN, or name). The provided text is what will be searched and redacted regardless of context.

  • Custom Regular Expression: Redacts all text that matches the Regex provided by the user. This option can be used to create partial redactions (e.g. redacting all but the last 4 digits of an SSN or Account Number), implement advanced "redact everything except..." rules, or create a basic Regex not yet available as a built-in preset.

  • SSN (###-##-####): This preset will match and redact all instances of the pattern ###-##-#### (i.e. 3 digits, a hyphen, 2 digits, a hyphen, 4 digits).

  • Date of Birth (mm/dd/yyyy): This preset will match instances where the Month and/or Day are one or two digits, and the Year is two or four digits.

  • Email Address (email@domain.com): This preset will match and redact email addresses which may or may not include numbers, special characters, and domains ending in any two or three characters (e.g. .us; .org; .gov; .edu; etc.).

  • Bank Account Number (10 or 12 digits): Redacts all text where the pattern matches ########## or ############ (i.e. 10 or 12 consecutive digits).

  • VIN (17 digit Alphanumeric): Redacts all text that matches the 17 character alphanumeric pattern associated with a VIN. As such, this preset may not identify VINs associated with vehicles manufactured prior to 1981.

Add regex or text to match - If selecting a preset the regex will automatically populate. If selecting a custom option, then enter the text or Regex in this field.

Rule Title - This is automatically populated based on the selected preset or custom option. The Rule Title is what will appear in the Filters as well as the Redaction Editor, so we strongly suggest users update it as needed so there's no misunderstanding as to what the rule is.

Redaction Label - This is the text that will appear over the redaction. "Redacted" is the default setting, but this can be customized as needed.


  1. Click OK to add the rule to the Auto Redaction job.

  2. To modify a rule, click somewhere on the rule to open the Add or Modify Rule window, make the desired changes, and click OK.

  3. To delete a rule, click the trashcan icon to the right of the rule you wish to delete.

Adding and Modifying Rules in an Auto Redaction Job

Processing an Auto Redaction Job

To process the Auto Redaction job, click "Run" which is located next to the job status in the job card. This will initiate processing which will update the job status to "Processing." As the job processes, the number of documents redacted as well as the number of documents remaining will automatically update as each document is searched and redacted.

Processing Progression Updates

Viewing Auto Redaction Job Results

Once the Auto Redaction job finishes processing, the job status will change to "Completed", and job totals for the number of documents redacted and the number of redactions made will be displayed in the job card. These figures are also available for each rule, and are presented as a bar graph with an alternative table view available. To switch your view, click the hamburger icon (the three horizontal lines) in the upper right corner.

Documents redacted as part of an Auto Redaction job can be viewed on a per rule, or per job basis as follows:

Viewing Redactions Created by a Specific Rule

Table View - Click on the hyperlinked rule, and all documents with a corresponding redaction will be populated on the Browse page.

Bar Graph View - Click on the burgundy Redactions bar for a specific rule, and all documents with a corresponding redaction will be populated on the Browse page.


Viewing All Redactions Created by a Job

In both views, you can access all redacted documents resulting from an Auto Redaction job by clicking on the icon shown below.

Click this Icon to View all Documents Redacted by this Job

Working With Auto Redactions in the Redaction Editor

Auto redacted documents can be viewed and edited in the Redaction Editor like any other document. In the Redaction Editor, redactions made via Auto Redaction will have an A next to them. Redactions made manually in the editor are denoted by an M. Additionally, if you hover of the A next to a redaction, the rule that generated that redaction will appear as shown in the image below.

Auto Redactions in the Redaction Editor

Mass Deleting Auto Redactions

Redactions created by an Auto Redaction rule can be removed simultaneously from all documents to which they were applied. To do this, access the Auto Redaction page and click on the appropriate Auto Redaction job. Identify the rule associated with the redactions you wish to remove, and then click the trashcan icon to initiate the removal process. A dialogue window will open asking you to confirm the deletion. Click OK to proceed.

At this time, deleting an Auto Redaction job will not remove all redactions associated with it. As stated above, redactions can only be removed by deleting the corresponding rule, or deleting them while in the Redaction Editor.

Rule Deletion Confirmation Message

If you require additional assistance, please contact Professional Services at ProfessionalServices@lexbe.com.