Attribute Changes Between Two Activities

Overview

The Attribute Changes between two Activities enrichment analyzes event attributes to detect changes that occur between two specific activities within each case. This powerful enrichment creates boolean attributes that indicate whether selected event attributes have changed in value between your specified activities. This is essential for understanding data transformations, handoff quality, process consistency, and identifying where information modifications occur in your business processes. The enrichment can analyze multiple event attributes simultaneously and even create new activities to mark where changes occurred, providing comprehensive visibility into data evolution throughout your process flow.

Beyond simple change detection, this enrichment helps organizations identify process inefficiencies, data quality issues, and compliance risks. By tracking attribute changes between key process milestones, you can pinpoint where data corrections happen, validate that required transformations occur, and ensure information consistency across departmental handoffs. The ability to generate activities at change points makes this enrichment particularly valuable for creating visual markers in process maps that highlight critical data transitions.

Common Uses

• Invoice Processing: Detect changes in invoice amounts, payment terms, or approval codes between submission and approval stages
• Order Management: Track modifications to order quantities, delivery dates, or customer requirements between order entry and fulfillment
• Change Request Handling: Monitor status changes, priority adjustments, or assigned team modifications between initial request and implementation
• Quality Control: Identify product specifications, test results, or quality scores that change between inspection stages
• Customer Service: Detect ticket priority changes, category reassignments, or resolution code modifications between creation and closure
• Healthcare Pathways: Track diagnosis code changes, treatment plan modifications, or insurance status updates between patient encounters
• Loan Processing: Monitor credit score updates, collateral valuations, or interest rate adjustments between application and approval

Settings

Event Columns: Select which event attributes to analyze for changes between the two activities. The enrichment will create a separate boolean attribute for each selected column, indicating whether its value changed. You can select multiple columns to comprehensively track data modifications across different aspects of your process. Leave empty to automatically analyze all non-system event attributes.

Activity 1: Choose the first activity that serves as the starting point for comparison. This activity marks where the initial attribute values will be captured. Select an activity that represents a meaningful checkpoint in your process where data should be in a specific state.

Activity 1 Selection Type: Specify whether to use the First or Last occurrence of Activity 1 within each case:

• First: Uses the earliest occurrence when the activity appears multiple times
• Last: Uses the most recent occurrence before Activity 2
• Default is First

Activity 2: Choose the second activity that serves as the ending point for comparison. The enrichment will compare attribute values at this activity against the values from Activity 1. Select an activity that represents another key milestone where you want to verify data changes or consistency.

Activity 2 Selection Type: Specify whether to use the First or Last occurrence of Activity 2 within each case:

• First: Uses the earliest occurrence after Activity 1
• Last: Uses the final occurrence in the case
• Default is Last

Create Activities: Enable this option to automatically inject new activities into your event log at points where attribute changes are detected. When enabled, the enrichment creates new events with activity names matching the changed attribute (e.g., "Invoice Amount-Change"). This provides visual markers in process maps and enables additional analysis of change patterns. Default is disabled.

Examples

Example 1: Invoice Amount Verification

Scenario: A finance team needs to track whether invoice amounts are modified between initial submission and final approval, as such changes require additional review according to company policy.

Settings:

• Event Columns: [Invoice_Amount, Tax_Amount, Total_Due]
• Activity 1: Submit Invoice
• Activity 1 Selection Type: First
• Activity 2: Approve Invoice
• Activity 2 Selection Type: Last
• Create Activities: False

Output: Creates three new boolean case attributes:

• Invoice_Amount-Change: True for cases where the invoice amount was modified
• Tax_Amount-Change: True for cases where tax calculations changed
• Total_Due-Change: True for cases where the total amount due changed

Sample data showing enrichment results: | Case ID | Invoice_Amount-Change | Tax_Amount-Change | Total_Due-Change | |---------|----------------------|-------------------|------------------| | INV-001 | False | False | False | | INV-002 | True | True | True | | INV-003 | False | True | True | | INV-004 | True | False | True |

Insights: The finance team discovers that 23% of invoices have amount changes between submission and approval, indicating a need for better initial validation. They implement additional training and system checks at the submission stage to reduce rework.

Example 2: Order Fulfillment Quality

Scenario: A logistics company wants to identify orders where delivery details change between order placement and shipment preparation, as these changes often lead to delivery delays and customer complaints.

Settings:

• Event Columns: [Delivery_Address, Delivery_Date, Shipping_Method, Order_Priority]
• Activity 1: Place Order
• Activity 1 Selection Type: First
• Activity 2: Prepare Shipment
• Activity 2 Selection Type: First
• Create Activities: True

Output: Creates four boolean attributes and injects new activities for each detected change:

• Delivery_Address-Change: Indicates address modifications
• Delivery_Date-Change: Shows delivery date adjustments
• Shipping_Method-Change: Reveals shipping method changes
• Order_Priority-Change: Tracks priority modifications

When changes are detected, new events are added to the log: | Case ID | Activity | Timestamp | |---------|----------|-----------| | ORD-123 | Place Order | 2024-01-10 09:00 | | ORD-123 | Delivery_Date-Change | 2024-01-10 14:30 | | ORD-123 | Shipping_Method-Change | 2024-01-10 14:30 | | ORD-123 | Prepare Shipment | 2024-01-10 14:30 |

Insights: Analysis reveals that 35% of orders have delivery date changes, primarily occurring during peak seasons. The company implements a customer notification system for date changes and adjusts capacity planning to reduce modifications.

Example 3: Healthcare Treatment Pathway Monitoring

Scenario: A hospital needs to track changes in patient diagnosis codes and treatment plans between initial emergency room assessment and admission to specialized departments, ensuring proper handoff communication.

Settings:

• Event Columns: [Diagnosis_Code, Treatment_Priority, Assigned_Department, Insurance_Status]
• Activity 1: ER Assessment
• Activity 1 Selection Type: First
• Activity 2: Department Admission
• Activity 2 Selection Type: First
• Create Activities: False

Output: Creates four boolean attributes for tracking medical data changes:

• Diagnosis_Code-Change: True when diagnosis is refined or changed
• Treatment_Priority-Change: Indicates priority level modifications
• Assigned_Department-Change: Shows department reassignments
• Insurance_Status-Change: Tracks insurance verification updates

Results enable filtering and analysis: | Case ID | Diagnosis_Code-Change | Treatment_Priority-Change | Assigned_Department-Change | |---------|----------------------|--------------------------|----------------------------| | PT-001 | True | False | False | | PT-002 | True | True | True | | PT-003 | False | False | False | | PT-004 | True | False | True |

Insights: The hospital identifies that 42% of patients have diagnosis code changes between ER and admission, indicating the need for better initial assessment protocols. They implement additional diagnostic tools in the ER to improve accuracy.

Example 4: IT Change Request Management

Scenario: An IT service desk wants to monitor how change request attributes evolve between initial submission and implementation start, identifying patterns that correlate with successful deployments.

Settings:

• Event Columns: [Risk_Level, Implementation_Type, Affected_Systems, Approval_Status]
• Activity 1: Submit Change Request
• Activity 1 Selection Type: First
• Activity 2: Start Implementation
• Activity 2 Selection Type: First
• Create Activities: True

Output: Creates boolean change indicators and activity markers:

• Risk_Level-Change: Indicates risk assessment modifications
• Implementation_Type-Change: Shows changes in implementation approach
• Affected_Systems-Change: Tracks scope modifications
• Approval_Status-Change: Monitors approval level changes

The enrichment injects activities to mark significant changes, enabling process mining visualization of where modifications occur in the change management workflow.

Insights: The IT team discovers that change requests with Risk_Level modifications have 3x higher failure rates. They implement mandatory review meetings when risk levels change to ensure proper planning adjustments.

Example 5: Manufacturing Quality Control

Scenario: A manufacturer needs to detect whether product specifications or quality measurements change between different inspection stations to identify where defects are introduced or corrected.

Settings:

• Event Columns: [Product_Weight, Color_Code, Quality_Score, Defect_Count]
• Activity 1: Initial Inspection
• Activity 1 Selection Type: Last
• Activity 2: Final Inspection
• Activity 2 Selection Type: Last
• Create Activities: False

Output: Creates quality change tracking attributes:

• Product_Weight-Change: Detects weight variations
• Color_Code-Change: Identifies color specification changes
• Quality_Score-Change: Tracks quality rating modifications
• Defect_Count-Change: Shows defect count changes

Analysis results by production line: | Production Line | % Weight Changes | % Quality Score Changes | % Defect Changes | |----------------|------------------|------------------------|------------------| | Line A | 2.3% | 15.2% | 18.5% | | Line B | 5.1% | 22.7% | 31.2% | | Line C | 1.8% | 8.9% | 11.3% |

Insights: Production Line B shows significantly higher change rates, indicating equipment calibration issues. The manufacturer schedules immediate maintenance and implements more frequent quality checks on that line.

Output

The enrichment creates new boolean case attributes for each selected event column, following the naming pattern [ColumnName]-Change. These attributes contain:

• True: When the event attribute value differs between Activity 1 and Activity 2
• False: When the event attribute value remains the same or when either activity is missing from the case
• Empty/Null: When the attribute cannot be evaluated (missing activities or attribute values)

Each created attribute:

• Is immediately available for use in filters, calculators, and other enrichments
• Can be exported with your enhanced dataset
• Appears in case attribute lists for analysis and visualization
• Supports process mining visualizations when change activities are created

When "Create Activities" is enabled, the enrichment also:

• Injects new events into the event log at the timestamp of Activity 2
• Names these events using the pattern [ColumnName]-Change
• Copies all other event attributes from Activity 2 to maintain context
• Requires dataset refresh to see new activities in process maps

The enrichment intelligently handles:

• Cases where one or both activities don't exist (no change recorded)
• Multiple occurrences of activities (controlled by selection type settings)
• Null or empty attribute values (treated as distinct values for comparison)
• Mixed data types (compares string representations)

Use these outputs to:

• Filter cases with specific types of changes for detailed analysis
• Calculate change rates and patterns across your process
• Create alerts for unexpected modifications
• Build conformance rules around allowed and prohibited changes
• Visualize change patterns in process maps when activities are created

---

This documentation is part of the mindzie Studio process mining platform.