Video: Find Obsolete Files Using Cognitive Suite

By John Bald

Watch: How to use Lucene query syntax to search files by create, modified and access dates and to tag them as obsolete

How to Find Obsolete Files

Step 1: Know the Syntax 

In order to find obsolete files you need to be able to search based on a date range using the Lucene query syntax. Below are the three Lucene equivalents to the standard Windows OS date metadata;

created date = creationTimeUtc:
modified date = lastWriteTimeUtc:
accessed date = lastAccessTimeUtc:

Lucene query examples:
extension:log AND creationTimeUtc:[2019-08-31 TO 2019-09-30]
extension:log AND creationTimeUtc:[* TO 2019-09-30]

Step 2: Determine Risk Levels

Risk levels might be unique to your organization and are completely customizable. Shinydocs has built a library of common rules that you can use as a starting point. 

Examples of Risk Levels:

Low Risk

.log files with a created date older than 6 months ago

Medium Risk

files with a modified date older than one year and contains the word "draft" in the file name or folder name

High Risk

files with a created date and modified date older than seven years


Step 3: Query the Visualizer Examples:

This is an optional step, but the interactive Visualizer tool makes the results of your query more easy to explore. 

search results

Step 4: Apply classifications in eDiscovery 

  1. In Discovery Desktop enter the same search string, which in this example returns a matching 23 files
  2. Right-click and select 'Select All' files 
  3. Right-click again and select 'Classification'
  4. Set the Classifications property to: 'Rot_obsolete'
  5. Enter the new classification value to: 'Low_risk_obsolete' 

Step 5: 

Repeat the steps above for both Medium and High Risk queries. That's it, machine speed at your fingertips!

How to Find Obsolete Files-00_03_13_00-Still002


Tags: Records Management, ROT, Obsolete, Videos, Cognitive Suite

Download a Sample Shinydocs Metadata Report

Get actionable recommendations to understand your unstructured data and start moving information from liability to viability.