Overview
The Group Attribute Values enrichment allows you to group multiple values of an existing attribute into categories, creating a new attribute with simplified values. This powerful data transformation tool helps you consolidate related attribute values into meaningful business categories, reducing complexity and improving analysis clarity. Instead of working with dozens or hundreds of unique values, you can create logical groupings that align with your business perspective.
This enrichment is particularly valuable when dealing with attributes that have many distinct values that can be logically combined. For example, you might group hundreds of detailed error codes into categories like "System Errors," "User Errors," and "Network Errors," or consolidate multiple payment methods into "Digital Payments" and "Traditional Payments." The enrichment uses filtering criteria to identify which cases or events should belong to each group, providing precise control over the categorization logic.
Common Uses
- Simplify complex categorizations: Group hundreds of product SKUs into product families or categories for clearer analysis
- Create business-relevant segments: Combine multiple customer types into strategic segments like "High Value," "Regular," and "New"
- Standardize regional variations: Group similar activities or statuses that vary by location into consistent categories
- Build performance indicators: Create binary attributes to identify cases meeting specific criteria (e.g., "Priority Customer" = True/False)
- Consolidate error types: Group detailed technical error codes into business-understandable categories
- Support decision making: Create simplified attributes for use in dashboards and reports that executives can understand
- Enable comparative analysis: Group cases into cohorts for before/after comparisons or A/B testing scenarios
Settings
Filter: Define the criteria that determine which cases or events belong to this group. You can use any combination of filters available in mindzieStudio, including attribute values, activity presence, date ranges, and complex logical conditions. The filter acts as the selection mechanism - all cases or events matching the filter will receive the group value.
New Attribute Name: Specify the name for the new attribute that will be created. This attribute will contain either the group names (for text grouping) or True/False values (for boolean grouping). Choose a descriptive name that clearly indicates the purpose of the grouping, such as "Customer Segment," "Error Category," or "Priority Case."
Boolean Group: When checked, creates a True/False attribute where cases matching the filter receive "True" and all others receive "False." This is ideal for binary classifications like "High Priority" (True/False) or "Requires Review" (True/False). When unchecked, you can specify a custom group name, allowing multiple groups to be created with different enrichment instances.
Group Name: (Only available when Boolean Group is unchecked) The text value to assign to cases or events that match the filter criteria. This allows you to create named categories like "Premium," "Standard," or "Basic." Multiple enrichments can target the same attribute name with different group names to build multi-category classifications.
Create Event Attribute: When checked, the enrichment creates an event-level attribute, evaluating the filter for each event individually. When unchecked (default), it creates a case-level attribute, evaluating the filter once per case. Use event attributes when the grouping logic depends on individual event characteristics rather than overall case properties.
Examples
Example 1: Customer Segmentation in Order Processing
Scenario: An e-commerce company wants to segment customers into "VIP," "Regular," and "New" categories based on order history and value for differentiated service levels.
Settings:
- Filter: Cases with Attribute "Total Order Value" > $10,000 AND "Order Count" > 20
- New Attribute Name: Customer Segment
- Boolean Group: Unchecked
- Group Name: VIP
- Create Event Attribute: Unchecked
Output: The enrichment creates a "Customer Segment" case attribute. Cases meeting the VIP criteria receive "VIP" as the value. Run additional enrichments with different filters and group names ("Regular" for medium values, "New" for first-time customers) targeting the same attribute name to complete the segmentation.
| Case ID | Total Order Value | Order Count | Customer Segment |
|---|---|---|---|
| C-001 | $15,000 | 25 | VIP |
| C-002 | $2,000 | 5 | Regular |
| C-003 | $500 | 1 | New |
Insights: The segmentation enables targeted analysis of process performance by customer tier, revealing that VIP customers experience 50% faster order processing but have more complex return processes requiring specialized handling.
Example 2: Manufacturing Quality Control Classification
Scenario: A manufacturing plant needs to identify production batches requiring quality review based on multiple sensor readings and inspection results exceeding thresholds.
Settings:
- Filter: Cases with Attribute "Temperature Variance" > 5 OR "Pressure Reading" > 100 OR "Visual Inspection" = "Failed"
- New Attribute Name: Requires Quality Review
- Boolean Group: Checked
- Create Event Attribute: Unchecked
Output: Creates a boolean "Requires Quality Review" attribute at the case level:
| Batch ID | Temperature Variance | Pressure Reading | Visual Inspection | Requires Quality Review |
|---|---|---|---|---|
| B-1001 | 3 | 95 | Passed | False |
| B-1002 | 7 | 98 | Passed | True |
| B-1003 | 2 | 105 | Passed | True |
| B-1004 | 4 | 90 | Failed | True |
Insights: Analysis shows 23% of batches require quality review, with temperature variance being the most common trigger. These batches have 3x longer cycle times due to additional inspection steps.
Example 3: Healthcare Patient Risk Categorization
Scenario: A hospital wants to categorize emergency room patients into risk levels based on symptoms and vital signs to optimize triage and resource allocation.
Settings:
- Filter: Cases with Attribute "Heart Rate" > 120 OR "Systolic BP" < 90 OR "Oxygen Saturation" < 92
- New Attribute Name: Patient Risk Level
- Boolean Group: Unchecked
- Group Name: High Risk
- Create Event Attribute: Unchecked
Output: Creates a "Patient Risk Level" attribute with "High Risk" for matching cases. Additional enrichments would define "Medium Risk" and "Low Risk" categories:
| Patient ID | Heart Rate | Systolic BP | O2 Saturation | Patient Risk Level |
|---|---|---|---|---|
| P-501 | 125 | 110 | 95 | High Risk |
| P-502 | 75 | 120 | 98 | Low Risk |
| P-503 | 90 | 85 | 94 | High Risk |
Insights: High-risk patients are immediately routed to critical care, reducing adverse events by 40%. Process mining reveals these patients have dedicated fast-track workflows with average door-to-treatment times under 10 minutes.
Example 4: Financial Transaction Fraud Indicators
Scenario: A bank needs to flag potentially fraudulent transactions based on unusual patterns in transaction attributes and customer behavior at the event level.
Settings:
- Filter: Events with Attribute "Transaction Amount" > $5,000 AND "Location Country" != "Home Country" AND "Time Since Last Transaction" < 60 seconds
- New Attribute Name: Potential Fraud Flag
- Boolean Group: Checked
- Create Event Attribute: Checked
Output: Creates an event-level boolean attribute marking individual transactions:
| Transaction ID | Amount | Location | Time Gap | Potential Fraud Flag |
|---|---|---|---|---|
| T-8001 | $7,500 | Foreign | 45 sec | True |
| T-8002 | $200 | Home | 2 hours | False |
| T-8003 | $5,100 | Foreign | 30 sec | True |
Insights: Transactions flagged as potential fraud trigger immediate review workflows. Analysis shows 85% accuracy in identifying actual fraud cases, with flagged transactions receiving additional authentication steps within 2 minutes.
Example 5: IT Incident Priority Grouping
Scenario: An IT service desk wants to consolidate dozens of incident subcategories into manageable priority groups for resource allocation and SLA management.
Settings:
- Filter: Cases with Attribute "Incident Type" IN ["Server Down," "Database Corrupt," "Network Outage," "Security Breach"]
- New Attribute Name: Incident Priority Group
- Boolean Group: Unchecked
- Group Name: Critical Infrastructure
- Create Event Attribute: Unchecked
Output: Consolidates multiple technical incident types into business-relevant groups:
| Incident ID | Incident Type | Original Priority | Incident Priority Group |
|---|---|---|---|
| I-901 | Server Down | P1 | Critical Infrastructure |
| I-902 | Password Reset | P3 | User Support |
| I-903 | Database Corrupt | P1 | Critical Infrastructure |
| I-904 | Software Install | P4 | User Support |
Insights: Critical Infrastructure incidents represent 15% of volume but consume 60% of senior technician time. These incidents follow expedited workflows with average resolution times of 2 hours versus 8 hours for standard issues.
Output
The Group Attribute Values enrichment creates a new attribute in your dataset with the following characteristics:
Attribute Type: The enrichment creates either a case attribute (default) or an event attribute based on the "Create Event Attribute" setting. Case attributes appear once per case and are visible in case-level analyses, while event attributes can vary across events within the same case.
Data Type: Boolean (when "Boolean Group" is checked) displaying as True/False, or String (when using custom group names) containing the specified text values.
Value Assignment: Cases or events matching the filter criteria receive either "True" (for boolean groups) or the specified group name (for text groups). Non-matching items receive "False" for boolean groups or retain their existing value/null for text groups.
Multiple Groups: You can create multiple enrichments targeting the same attribute name with different filters and group names. This builds multi-category classifications where each case receives the appropriate category based on which filter it matches. If a case matches multiple filters, the last-applied enrichment takes precedence.
Integration: The new grouped attribute integrates seamlessly with all mindzieStudio features including filters, calculators, and visualizations. Use these simplified attributes in process maps to show flow variations by group, in dashboards for comparative metrics, or as filter criteria in other enrichments.
This documentation is part of the mindzie Studio process mining platform.