Table – Profiling

The Profiling screen provides a detailed overview of your table's structure and content. It helps identify potential issues and supports rule creation by showing statistics and relationships between columns.

(1) Summary statistics

The summary statistics section shows an overview of each column in the table based on a sample of up to 250,000 rows. For each column, the following metrics are displayed:

• NULL – Share of missing values

• UNIQUE – Share of unique values

• MIN/MAX/AVERAGE – Minimum, maximum, and average values (for numeric/date fields)

• PII – Highlights potential Personally Identifiable Information

PII refers to data that can be used to identify an individual. Because of legal and compliance requirements, PII often requires special handling.

We scan for a broad range of PII categories using a combination of machine learning and regex-based detection. Categories are assigned probabilistically and may not always be perfectly accurate.

PII categories include:

• Social Security Number / US SSN

• German Passport Number

• Credit Card Number

• IBAN

• Person Name

• Email

• Phone Number

• IP Address

• URL

(2) Quality status

Displays the overall quality score and summary for the table based on the most recent check.

• Score – Percentage of quality based on all active rules

• Issues – Total number of current data quality issues

• Table check scope – Scope of validation (e.g., full table)

• Rows / columns – Total entries and attributes checked

• Active rules – Number of rules currently active

(3) Table-based rules

Shows the current status of table-level rules like:

• Format – Verifies structural consistency (e.g., expected pattern)

• Freshness – Checks how recently the data has been updated

• Volume – Monitors row count for major changes

(4) DQ-AI Assistant

You can use the DQ-AI Assistant to:

• Ask questions about your data

• Generate new rule suggestions

• Get help with data interpretation or insights

(5) Correlation matrix

The correlation matrix visualizes relationships between different columns in the dataset.

• Yellow = Positive correlation

• Blue = Negative correlation

• Hover over any intersection to see the exact correlation value

Notes

• Profiling helps you decide where to apply the most relevant rules

• Statistics are based on a maximum of 250,000 sample rows

• Correlation matrix is useful for identifying redundant or dependent fields

• Learn more: Table – Rules Screen, Table – Issue Screen, Table-Based Rules Overview