/sap-datasphere

SAP Datasphere Catalog and Governance

Overview

The SAP Datasphere Catalog provides a centralized hub for discovering, understanding, and governing data assets. It enables organizations to implement data governance practices while making trusted data accessible to business users.

Documentation: https://help.sap.com/docs/SAP_DATASPHERE/aca3ccb4b2f84eb8b6154e8fd2812c0e

Catalog Features

Asset Discovery

The catalog allows users to discover and explore:

• Data Products: Curated packages of views and models for specific use cases
• Views and Tables: Individual data objects with metadata
• Analytic Models: Consumer-ready analytics with measures and dimensions
• Glossary Terms: Business definitions and context
• KPIs: Key performance indicators with calculations

Catalog Capabilities

Glossary Management

Creating a Glossary

A business glossary provides standardized definitions for key business terms.

Glossary Structure:
  Categories:
    - Finance
    - Sales
    - HR
    - Supply Chain

  Term Components:
    - Name: Business-friendly term
    - Definition: Clear, unambiguous description
    - Examples: Usage examples
    - Synonyms: Alternative names
    - Related Terms: Links to related definitions
    - Owner: Responsible party
    - Status: Draft, Approved, Deprecated

Example Glossary Terms

Revenue

Term: Revenue
Category: Finance
Definition: The total income generated from sales of goods or services before any expenses are deducted
Calculation: SUM(Net Sales) for a given period
Synonyms: Sales, Turnover, Income
Related Terms: Gross Revenue, Net Revenue, Deferred Revenue
Owner: Finance Data Steward
Status: Approved

Customer Lifetime Value (CLV)

Term: Customer Lifetime Value
Abbreviation: CLV
Category: Sales
Definition: The predicted net profit attributed to the entire future relationship with a customer
Calculation: (Average Order Value × Purchase Frequency × Customer Lifespan) - Acquisition Cost
Related Terms: Customer Acquisition Cost, Churn Rate
Owner: Marketing Analytics
Status: Approved

Linking Terms to Data Assets

Associate glossary terms with views and columns:

1. Navigate to view in Data Builder
2. Open Business Purpose section
3. Link relevant glossary terms to:
   • The view itself (overall purpose)
   • Individual columns (column meaning)

Data Quality

Quality Rules

Define rules to validate data quality:

Quality Rule Types:
  Completeness:
    - Not Null checks
    - Required field validation

  Uniqueness:
    - Primary key uniqueness
    - Duplicate detection

  Validity:
    - Range checks (min/max)
    - Pattern matching (regex)
    - Domain values (allowed list)

  Timeliness:
    - Freshness checks
    - SLA monitoring

  Consistency:
    - Cross-field validation
    - Referential integrity

Quality Score Calculation

Quality Score = (Passed Records / Total Records) × 100

Aggregate Score = Weighted Average of:
  - Completeness Score (weight: 0.25)
  - Uniqueness Score (weight: 0.25)
  - Validity Score (weight: 0.30)
  - Timeliness Score (weight: 0.20)

Quality Monitoring

Set up continuous quality monitoring:

1. Create Quality Rules: Define expectations
2. Schedule Validation: Run rules on schedule
3. Set Thresholds: Define acceptable quality levels
4. Configure Alerts: Notify when quality drops
5. Track Trends: Monitor quality over time

Data Classification

Sensitivity Levels

Classify data by sensitivity:

Data Categories

Auto-Classification

Configure automatic classification based on:

1. Column Names: Match patterns like *_SSN, *_EMAIL
2. Data Patterns: Detect formats like phone numbers, credit cards
3. Source Systems: Apply rules based on origin
4. Glossary Terms: Inherit classification from linked terms

Data Lineage

Lineage Visualization

View data flow from source to consumption:

Source System → Remote Table → Staging View → Fact View → Analytic Model → SAC Story

Lineage Information

Impact Analysis

Before making changes, understand impact:

1. Select object in Data Builder
2. Click Impact Analysis
3. Review:
   • Downstream dependencies
   • Affected reports/stories
   • User impact

Publishing to Catalog

Publication Workflow

1. Create Object → 2. Add Metadata → 3. Link Terms → 4. Request Approval → 5. Publish

Required Metadata for Publication

Publication Requirements:
  Required:
    - Business Name (readable name)
    - Description (purpose and content)
    - Owner (responsible person)
    - Classification (sensitivity level)

  Recommended:
    - Glossary Term Links
    - Quality Score
    - Data Freshness
    - Sample Data
    - Usage Guidelines

Approval Process

1. Submit for Review: Owner submits asset
2. Steward Review: Data steward validates
3. Quality Check: Automated quality validation
4. Approve/Reject: Decision with feedback
5. Publish: Make available in catalog

Governance Roles

Standard Roles

Role Assignments

Configure governance roles in Administration:

Governance Role Assignment:
  Data Steward - Finance:
    User: finance.steward@company.com
    Scope: Finance domain objects
    Permissions:
      - Approve publications
      - Edit glossary terms
      - Define quality rules

Policies and Compliance

Data Retention Policies

Retention Policy - Transaction Data:
  Category: Financial Transactions
  Retention Period: 7 years
  Archive After: 2 years
  Delete After: 7 years
  Legal Basis: Tax regulations

Retention Policy - Log Data:
  Category: System Logs
  Retention Period: 90 days
  Archive After: 30 days
  Delete After: 90 days

Access Policies

Combine with Data Access Controls:

Access Policy - Regional Data:
  Rule: Users can only access data from their assigned region
  Implementation:
    - Data Access Control: region_access
    - Permission Table: user_region_assignments
    - Criteria Column: REGION_ID

Audit Logging

Track all governance activities:

Best Practices

Governance Implementation

1. Start Small: Begin with critical data domains
2. Executive Sponsorship: Secure leadership support
3. Clear Ownership: Assign accountable owners
4. Automated Monitoring: Don't rely on manual checks
5. Regular Reviews: Audit governance effectiveness quarterly

Catalog Population

1. Prioritize by Usage: Catalog most-used assets first
2. Quality over Quantity: Well-documented few > poorly-documented many
3. Template Descriptions: Create standard description templates
4. Glossary First: Build glossary before linking to assets

User Adoption

1. Training: Educate users on catalog search
2. Quick Wins: Show value with popular datasets
3. Feedback Loop: Collect and act on user feedback
4. Gamification: Recognize top contributors

Integration with SAP Analytics Cloud

Catalog-Enabled Stories

Users can:

1. Search catalog from SAC
2. Preview data before adding
3. See lineage and quality info
4. Add trusted data to stories

Trusted Data Badge

Assets meeting criteria receive "Trusted" badge:

• Published to catalog
• Quality score > 90%
• Complete documentation
• Active owner assigned