Searching in Relativity Flashcards
dtSearch
Advanced search functionality - proximity, stemming & fuzzy searches across field type
dtSearch - Creating a dtSearch Index
- Navigate to Search Indexes tab and click “New dtSearch Index”
- Complete the Fields
- Click Save to display Index Details page
- Click “Build Index: Full” -> Dialogue box to verify and select “Activate this index upon completion”
- Click “OK”; once complete, click “Activate Index” if you didn’t select “Activate this index upon completion”
- Click “Refresh Page” to see index’s current build status
dtIndex Fields
Name Order Searchable Set Email Notification Recipients Advanced Settings - Auto recognize date, email, credit card numbers - Create accent sensitive - sub-index size (minimum value is 1,000) - sub-index fragmentation threshold - Index Share Noise Words Alphabet
Using dtSearch syntax options
Boolean - AND, OR, NOT Wildcards - ?, * W/N Operator - W/N (or WI) Proximity with order - PRE Built-in search words - xfirstword, xlastword Operator precedence - () [parenthesis] Search Operator Words - use quotation marks Fuzzy Searching - % Stemming - ~ RegEx - "##" Noise Words Auto-recognition of dates, emails, credit cards - date(), mail(), creditcard() Phonic Searching - # Numerical Patterns - =
dtSearch Notes
- Underscore “_” is not recognized as a space by default
- Colon (:) and (&) are not used by Rel but still considered a syntax term by dtSearch – to search for these, must modify the alphabet file
- dtSearch indexes are case insensitive by default
- Searching for words right next to each other with no operator is an exact phrase search
- only need quotes when searching for an operator
dtSearch - Auto-recognition
- Date Recognition - searches for strings that appear to be dates
- date(january 10 2006)
- range: date(january 5 2006 to january 20 2006)
- Email address recognition - can search for email addresses regardless of alphabet settings for “@”, “.”, or other punctuation
- mail(s*@example.com)
- Credit Card Number recognition - any sequence that matches the syntax
- creditcard(1234*)
Numerical Patterns
To search for numerical patterns such as social security numbers, you can use the “=” wildcard, which matches any single digit
- more performant than a regular expression for the same pattern
Searching Workflows - Workflow Details
Basic Keyword Search - populates automatically, supports basic Boolean (AND, OR, NOT, *)
dtSearch Index is custom built and must be set up with an index to query against; supports advanced functionality
Search - Best Practices for Advanced Operators - Proximity
- use pre /n for directional proximity, w /n for non-directional
- xfirstword or xlastword w /n to search beginning and end of doc
- distance between terms is important (1 w/1 2) w/2 3 to catch 1 2 3
- noise words count as words for proximity
- punctuality counts as whitespace under default settings
- line breaks and consecutive space characters are single spaces
- stop words and connector words count as words when calculating proximity
Troubleshooting Workflow
- use a white board approach to map out searches before you run them
Creating or Editing a Saved Search
1, Click the magnifying glass with doc icon
- Click “Create New Search” or “Edit”
- Fill out the information in the Information sectoin
- Click “Add Condition” or click “Add Logic Group”
- Click on the Fields tab, and choose which fields to display search results
- Click on Sort tab, and specify any sorting
- Click on the Other Tab to add additional info
- Click “Save & Search”
Search Fields - Information
Name
Owner
Dashboard
Requires Manual Rerun - for searches that might take a long time to run, and you don’t need them to run automatically when you navigate back
Scope - Entire Workspace / Selected Folders (include subfolders)
Notes
Search Fields - Advanced
Relativity Applications
Keywords
Query Hint - only use if instruction by the Relativity Client Services team
Search Fields - Conditions
Add Condition
Index Search [Keyword / dtSearch / Analytics]
Saved Search
Add Logic Group
Search Panel
Available in the Doc List and many other tabs
- click the magnifying glass to show
- can auto-run or not as new conditions are added
- can apply logic groups
- can copy conditions by clicking pop-up icon in corner
- can toggle conditions on and off
- can collapse a card or remove condition
Search Term Reports - Guidelines
- define a saved search for the Searchable Set
- related items will not be searched
- confirm dtSearch index includes all docs in searchable set
- Select “Tag Hits” to create a multiple object field (if this isn’t set can’t select the STR in the Highlight Fields option)
- Only terms in current reports will highlight; if you add new search terms to reports, you must run pending terms so they appear highlighted in documents
- System automatically preserves the precise order in which the terms were entered when generating a report
Creating a Search Term Report
- Go to Search Terms Report tab
- Click “New Search Terms Report”
- Complete Fields
- Click Save
Creating a Search Term Report - Fields
Name Index Searchable Set Include Relational Group Tag Hits Show in Field Tree Calculate Unique Hits Remove Hidden Characters (filters out hidden or non-displayable text control characters)
Searching with the Entity Object and Name Normalization - Emails w/in a particular org
Use the Alias From::Domain and Alias Recipient::Domain
NB - can also use to identify specific cross-org comms (i.e. with outside counsel)
- use double negative - Alias From -> NOT THESE CONDITIONS -> Alias From::Domain IS NOT [target domain]
Communications between 2 specific individuals and no others
(Entity From IS SET
AND
(Entity From NOT THESE CONDITIONS Entity From NONE OF THESE [Entity A]
AND
Entity Recipient NOT THESE CONDITIONS Entity Recipient NONE OF THESE [Entity B])
OR
To limit to these 2 and no one else, use “Email Recipient Count” set to 1
Entities that Communicated on Privileged Documents
- edit the Privilege Designation field to “Open to Associations” is “Yes”
Entity Participant THESE CONDITIONS Entity Participant::Privilege Designation ANY OF THESE
Emails where a specific person drops off a communication
Entity Participant -> Entity Participant ANY OF THESE [Entity]
AND
Entity From NOT THESE Entity From ANY OF THESE [Entity]
AND
Entity Recipient NOT THESE Entity Recipient ANY OF THESE [Entity]
Regular Expressions - Use Cases
Helpful where you need to find different numbers/letters that contain the same pattern