Normalize columns

Description

The Normalize Columns activity allows you to apply one or more normalization techniques to selected text columns within a dataset. This is useful for improving data quality and consistency across structured text inputs.

You can configure each column with its own normalization method, such as trimming spaces, standardizing capitalization, removing special characters, and more.

Use this activity to:

• Standardize inconsistent text formats (e.g., names, categories)
• Prepare data for matching, lookups, or deduplication
• Clean imported datasets for better presentation or downstream processing

Use case:
Normalize names, job titles, and category fields in user-uploaded datasets for consistency before mapping them to master data or applying classification rules.

Input

Input Type	Status
Data	Required

Output

Output Type	Format	Description
Data	Table	Data with normalized column values

Configuration Fields

Field Name	Description
Column Map	Specify columns and assign one or more normalization techniques to each.
Include Original	If enabled, both original and transformed columns are included in output. If disabled, only the normalized columns are returned.

Normalization Options

• Convert to lowercase – e.g., “John” → “john”
• Convert to uppercase – e.g., “John” → “JOHN”
• Convert to title case – e.g., “john doe” → “John Doe”
• Capitalize first letter – e.g., “software engineer” → “Software engineer”
• Trim whitespace – Removes leading and trailing spaces
• Remove whitespace – Deletes all whitespace characters
• Remove special characters – Strips out non-alphanumeric characters (e.g., &, %, #)
• Normalize accents – Replaces accented characters with their unaccented equivalents (e.g., “é” → “e”)

Sample Input

ID	Name	Description	Category
1	jOHN DOE	Software Engineer	IT & Tech
2	jane SMITH	Data Scientist	Analytics
3	mark_o’leary	Machine Learning	AI & ML

Sample Configuration

Column	Normalization Applied
Name	Title Case
Description	Uppercase
Category	Remove Special Characters, Title Case

Include Original: Enabled

Sample Output

ID	Name	Description	Category
1	John Doe	SOFTWARE ENGINEER	IT Tech
2	Jane Smith	DATA SCIENTIST	Analytics
3	Mark O’Leary	MACHINE LEARNING	AI ML