hide blank attributes

Overview

The Hide Blank Attributes enrichment is an automated data cleanup operator that identifies and removes all attributes (columns) that contain no data across your entire dataset. This powerful cleanup tool scans both case-level and event-level attributes, automatically hiding any columns where every single row contains null or empty values. By removing these empty columns, the enrichment significantly simplifies your dataset view, reduces visual clutter in analysis tools, and improves performance when working with large datasets.

This enrichment is particularly valuable when importing data from enterprise systems that export fixed schemas with many optional fields, or when working with datasets that have undergone multiple transformations where some attributes become obsolete. Unlike manual column removal which requires identifying each empty column individually, this enrichment performs a comprehensive sweep of your entire dataset in a single operation. The enrichment preserves all columns that contain at least one non-null value, ensuring no potentially useful data is lost while maximizing the cleanliness and usability of your process mining workspace.

Common Uses

Clean imported datasets from ERP systems that include hundreds of optional fields with no populated data
Simplify dataset views after filtering operations that may leave certain attributes completely empty
Reduce visual clutter in the case and event attribute panels to focus on meaningful data
Improve performance and reduce memory usage when working with wide datasets containing many unused columns
Prepare datasets for export or sharing by removing irrelevant empty columns
Clean up after data transformations that consolidate multiple attributes into new calculated fields
Streamline conformance checking by removing attributes that provide no analytical value

Settings

This enrichment operates automatically without requiring any configuration. It scans all non-calculated and non-hidden attributes in your dataset, removing only those that are completely empty across all cases and events.

Examples

Example 1: Cleaning ERP System Export

Scenario: A manufacturing company exports order processing data from SAP with over 200 standard fields, but their specific implementation only uses about 60 fields, leaving 140+ columns completely empty and making analysis difficult.

Before Enrichment: Dataset contains 215 total attributes including:

Case Attributes: 125 columns (75 empty)
Event Attributes: 90 columns (65 empty)
Examples of empty columns: Legacy_System_ID, Deprecated_Cost_Center, Old_Warehouse_Code, Custom_Field_1 through Custom_Field_50

After Enrichment: Dataset simplified to 75 meaningful attributes:

Case Attributes: 50 columns (all containing data)
Event Attributes: 25 columns (all containing data)
All empty columns automatically hidden from view

Output: The enrichment removed 140 empty columns while preserving all 75 columns that contained at least one value. The dataset view is now focused only on attributes with actual data, making navigation and analysis significantly easier.

Insights: After cleanup, analysts could quickly identify the relevant attributes for process mining. The simplified view revealed that order processing actually involved only 12 key attributes for decision-making, which were previously hidden among hundreds of empty fields. Performance improved by 40% when loading the dataset due to reduced memory overhead.

Example 2: Post-Filtering Cleanup in Healthcare

Scenario: A hospital filters their patient treatment dataset to analyze only emergency department cases, which causes many specialized ward attributes to become completely empty since emergency cases don't use those data fields.

Before Enrichment: After filtering to emergency cases only:

Total Attributes: 180
Populated Attributes: Emergency_Triage_Level, Emergency_Wait_Time, Emergency_Treatment
Empty Attributes: ICU_Ventilator_Settings, Surgery_Type, Rehabilitation_Plan, Oncology_Stage, and 85 other specialized department fields

After Enrichment:

Total Visible Attributes: 92
All attributes now contain relevant emergency department data
88 empty specialized department attributes hidden

Output: The enrichment automatically identified and hid all attributes that became empty after the emergency department filter was applied. The remaining attributes all contain data relevant to emergency cases.

Insights: The cleaned dataset allowed emergency department managers to focus on their specific KPIs without distraction from irrelevant fields. Analysis time decreased by 60% as staff no longer had to scroll through empty columns to find relevant data.

Example 3: Financial Process Consolidation

Scenario: A bank merges invoice processing data from three different systems, each with unique field structures, resulting in many system-specific attributes being empty for cases from other systems.

Before Enrichment: Merged dataset with 340 attributes:

Common fields (used by all systems): 45 attributes
System A specific fields: 95 attributes (empty for System B and C cases)
System B specific fields: 110 attributes (empty for System A and C cases)
System C specific fields: 90 attributes (empty for System A and B cases)

After Enrichment: Focused dataset with 45 common attributes visible, plus only the system-specific attributes that contain data for the current case selection.

Output: The enrichment removed all columns that were completely empty, leaving only the 45 common fields that all three systems populate. System-specific attributes that were empty across the entire merged dataset were automatically hidden.

Insights: The consolidation revealed that despite different system structures, all three systems captured the same 45 core process attributes. This discovery enabled the bank to standardize their invoice processing across all systems and reduce complexity by 85%.

Example 4: Procurement Data Preparation

Scenario: A retail company's procurement dataset includes attributes for various approval levels and special handling codes, but many of these fields are only used for high-value or regulated items, leaving them empty for routine purchases.

Before Enrichment: Dataset with 150 attributes including:

Standard fields: PO_Number, Supplier, Amount, Create_Date (always populated)
Conditional fields: VP_Approval, Legal_Review, Hazmat_Code, Export_License, Compliance_Check (95% empty)
Legacy fields: Old_Vendor_Code, Previous_System_Ref (100% empty after migration)

After Enrichment: Streamlined dataset with 67 active attributes:

All standard procurement fields retained
Conditional fields with at least one value retained
Completely empty legacy fields removed

Output: The enrichment hid 83 attributes that contained no data, including all legacy fields and conditional approval fields that were never used in the current dataset. The remaining attributes all contribute to process analysis.

Insights: After cleanup, the procurement team discovered that only 5% of purchases actually required special approvals, allowing them to streamline the process for the 95% of routine purchases. The simplified view made it easy to identify these high-complexity cases for separate analysis.

Example 5: Manufacturing Quality Control

Scenario: An automotive parts manufacturer exports quality control data with hundreds of measurement fields, but each production line only uses specific measurements relevant to their parts, leaving many fields empty.

Before Enrichment: Quality dataset with 450 attributes:

Common fields: Part_Number, Production_Line, Timestamp, Pass_Fail (always populated)
Line-specific measurements: 200+ measurement fields per line (empty for other lines)
Deprecated measurements: 50+ old quality metrics no longer collected

After Enrichment: Relevant dataset with 125 attributes:

All common fields preserved
Only measurements with data retained
All deprecated and unused measurement fields hidden

Output: The enrichment removed 325 empty measurement columns while preserving the 125 columns containing actual quality data. Each production line's view now shows only relevant measurements.

Insights: The cleanup revealed that despite having 450 possible measurements, each production line only actively monitored 20-30 critical quality metrics. This insight led to a focused quality improvement program that reduced defect rates by 15% by concentrating on the measurements that actually mattered.

Output

The Hide Blank Attributes enrichment modifies the visibility of existing columns without deleting data:

Hidden Columns:

Case attributes where every case row contains null/empty values
Event attributes where every event row contains null/empty values
Columns are marked as hidden but not deleted from the dataset
Hidden status can be reversed if needed through column management

Preserved Columns:

All columns containing at least one non-null value
All calculated columns (created by other enrichments)
All columns already marked as hidden (no redundant processing)
System columns like Case ID and Activity names

Performance Impact:

Reduced memory usage when loading datasets
Faster rendering of attribute lists and filters
Improved query performance on simplified column sets
Cleaner export files when sharing datasets

The enrichment's effects are immediately visible in the case and event attribute panels, where empty columns no longer appear. This creates a focused, efficient workspace for process analysis.