Overview
The Cases with Changed Attribute filter selects cases where an event attribute has different values across events within the case. This case-level filter examines all events in each case and keeps only those cases where the specified attribute varies from event to event. The filter is particularly useful for identifying dynamic processes where attribute values change during case execution, such as status transitions, location changes, or resource handoffs.
The filter focuses on non-null values - it filters out null values first, then checks if the remaining values are identical. Cases where all events have the same value (or all null values) are excluded from the results.
Common Uses
- Status Progression Analysis: Identify cases where status values changed during processing, indicating progression through different stages.
- Resource Handoff Detection: Find cases where different resources or departments handled different events, revealing collaboration patterns.
- Location Change Tracking: Discover cases where the location attribute changed, indicating physical movement or transfer between sites.
- Priority Escalation: Detect cases where priority levels changed during execution, showing escalation or de-escalation patterns.
- Process Deviation Identification: Find cases where attribute values varied unexpectedly, potentially indicating exceptions or non-standard processing.
- Multi-Stage Process Analysis: Identify cases that went through multiple stages by detecting changes in stage-related attributes.
Settings
Event Column Name: Select the event attribute you want to evaluate for variation. The filter will return cases where this attribute has different values across events. Only non-null values are considered when checking for variation.
Note: The attribute must exist in the event table and be of a supported data type (String, Int32, Int64, DateTime, TimeSpan, Single, Double, or Boolean). If you mistype the column name, the filter's validation system will suggest similar column names.
Examples
Example 1: Finding Cases with Status Changes
Scenario: You want to identify all order processing cases where the order status changed during processing, indicating that the order progressed through different stages.
Settings:
- Event Column Name: "Order Status"
Result: The filter returns cases where different events have different "Order Status" values (e.g., "New" -> "Processing" -> "Shipped"). Cases where all events have the same status are excluded.
Insights: This helps identify:
- Cases that successfully progressed through the workflow
- Normal processing patterns with status transitions
- Cases that may have experienced multiple status changes
- Orders that moved through different fulfillment stages
Example 2: Detecting Resource Handoffs
Scenario: You need to find cases where different resources or employees worked on different activities, indicating collaboration or handoff situations.
Settings:
- Event Column Name: "Resource"
Result: The filter selects cases where the "Resource" attribute varies across events, meaning multiple people or systems handled different activities.
Insights: These cases reveal:
- Collaborative work patterns where multiple people contribute
- Handoffs between departments or teams
- Cases requiring specialized expertise from different resources
- Potential bottlenecks where resource changes occurred
Example 3: Identifying Location Changes
Scenario: You want to track shipments or items that moved between different locations during processing.
Settings:
- Event Column Name: "Location"
Result: The filter returns cases where events occurred at different locations, indicating physical movement or transfer.
Insights: This can reveal:
- Items that traveled through multiple warehouses or distribution centers
- Cross-site processing patterns
- Geographic routing of cases
- Cases requiring multi-location coordination
Example 4: Finding Priority Escalations
Scenario: Identify support tickets or requests where the priority level changed during handling, indicating escalation or de-escalation.
Settings:
- Event Column Name: "Priority"
Result: The filter selects cases where "Priority" values changed between events (e.g., from "Low" to "High").
Insights: These cases might indicate:
- Escalated issues requiring increased attention
- De-escalated cases after initial assessment
- Dynamic priority adjustments based on customer feedback
- Cases requiring management intervention
Example 5: Detecting Department Transfers
Scenario: You want to find cases that were transferred between departments during processing, indicating complex cases requiring cross-functional support.
Settings:
- Event Column Name: "Department"
Result: The filter returns cases where different events were handled by different departments.
Insights: This helps identify:
- Cases requiring expertise from multiple departments
- Cross-functional collaboration patterns
- Potential handoff delays between departments
- Complex cases that could benefit from process optimization
Example 6: Tracking Approval Level Changes
Scenario: Identify cases where the approval level changed, such as requests that required escalation to higher management levels.
Settings:
- Event Column Name: "Approval Level"
Result: The filter selects cases where the "Approval Level" attribute varied across events.
Insights: These cases may represent:
- Requests that required multiple approval tiers
- Escalation patterns for high-value or complex requests
- Cases that exceeded standard approval thresholds
- Multi-stage approval workflows
Output
The filter returns a new dataset containing only the cases where the specified event attribute has different values across events. Each returned case preserves all its original events and attributes, and you will see variation in the selected attribute's values.
Cases where all non-null values are identical are excluded. Cases with only null values or a single unique value are also excluded.
If no cases match the criteria, the filter returns an empty result set.
Technical Notes
- Filter Type: Case-level filter (removes entire cases, not individual events)
- Null Handling: Ignores null values when checking for variation - only considers non-null values
- Variation Detection: Compares the first non-null value against all other non-null values
- Supported Data Types: String, Int32, Int64, DateTime, TimeSpan, Single, Double, Boolean
- Validation: Automatically suggests similar column names if the specified column is not found
This documentation is part of the mindzieStudio process mining platform.