🛡️ Tutorial 07: Governance & Purview¶

Last Updated: 2026-04-15 | Version: 2.0 Status: ✅ Final | Maintainer: Documentation Team

🏠 Home > 📖 Tutorials > 🛡️ Governance & Purview

🛡️ Tutorial 07: Governance & Purview - Data Catalog & Lineage¶

📊 Progress Tracker¶

+--------+--------+--------+--------+--------+--------+--------+--------+--------+--------+
|   00   |   01   |   02   |   03   |   04   |   05   |   06   |   07   |   08   |   09   |
| SETUP  | BRONZE | SILVER |  GOLD  |  RT    |  PBI   | PIPES  |  GOV   | MIRROR |  AI/ML |
+--------+--------+--------+--------+--------+--------+--------+--------+--------+--------+
                                                                    ^
                                                                    |
                                                               YOU ARE HERE

📖 Overview¶

This tutorial covers implementing data governance using Microsoft Purview integration with Fabric. You will catalog data assets, apply classifications, track data lineage, and establish compliance controls for the regulated casino/gaming environment.

Microsoft Purview Hub in Fabric¶

The Purview hub in Fabric provides centralized governance capabilities for your data estate:

Source: Use Microsoft Purview hub in Fabric

flowchart TB
    subgraph Fabric["Microsoft Fabric"]
        LH[Lakehouses]
        SM[Semantic Models]
        PBI[Power BI Reports]
    end

    subgraph Purview["Microsoft Purview"]
        DC[Data Catalog]
        DM[Data Map]
        CL[Classifications]
        GL[Glossary]
        LN[Lineage]
        POL[Policies]
    end

    Fabric <-->|Scan & Govern| Purview

    DC --> DM
    DM --> CL
    CL --> GL
    GL --> LN
    LN --> POL

🎯 Learning Objectives¶

By the end of this tutorial, you will be able to:

• Connect Microsoft Purview to Fabric workspace
• Scan and catalog all data assets
• Apply classifications and sensitivity labels
• Create business glossary terms for casino domain
• View and analyze data lineage
• Configure data access policies
• Generate compliance reports

⚠️ Why Governance Matters for Casino Data¶

Casino and gaming operations handle highly regulated data:

Warning: Failure to properly govern casino data can result in regulatory fines, license revocation, and criminal penalties.

📊 Microsoft Purview Capabilities¶

Microsoft Purview provides unified data governance:

Microsoft Purview enables a federated approach to data governance. Source: Data governance solutions

📋 Prerequisites¶

• Completed Tutorials 01-06 (Medallion architecture + Pipelines)
• Microsoft Purview account deployed in your Azure subscription
• Purview Data Curator role or higher
• Fabric workspace admin access
• Microsoft Entra ID permissions for service principal (optional)

Note: Purview is included with Microsoft 365 E5 or can be provisioned separately in Azure.

🛠️ Step 1: Connect Purview to Fabric¶

1.1 Register Fabric as a Data Source¶

1. Open Microsoft Purview Portal
2. Navigate to Data Map > Sources
3. Click Register > Microsoft Fabric
4. Configure registration:

1. Click Register

1.2 Configure Scan Credentials¶

1. Go to Management > Credentials
2. Create new credential:
3. Name: fabric-scan-credential
4. Authentication method: Managed Identity (recommended) or Service Principal
5. Grant credential access to Fabric workspace

1.3 Create Scan¶

1. On the registered source, click New Scan
2. Configure scan settings:

1. Scope your scan - Select:
2. lh_bronze
3. lh_silver
4. lh_gold
5. Semantic models

6. Reports

7. Scan rule set: System default (or create custom)

8. Click Continue

1.4 Set Scan Trigger¶

Configure recurring scans:

Click Save and Run to start initial scan.

sequenceDiagram
    participant P as Purview
    participant F as Fabric Workspace
    participant L as Lakehouses

    P->>F: Authenticate (Managed Identity)
    F-->>P: Token granted
    P->>L: Scan lh_bronze
    L-->>P: Schema + Metadata
    P->>L: Scan lh_silver
    L-->>P: Schema + Metadata
    P->>L: Scan lh_gold
    L-->>P: Schema + Metadata
    P->>P: Apply Classifications
    P->>P: Build Lineage Graph

🛠️ Step 2: Review Scanned Assets¶

2.1 View Data Catalog¶

1. Go to Data Catalog > Browse
2. Filter by:
3. Source type: Microsoft Fabric
4. Source: Fabric-Casino-POC

You should see: - Lakehouses (lh_bronze, lh_silver, lh_gold) - Delta tables within each lakehouse - Semantic models - Power BI reports

2.2 Explore Table Details¶

Click on a table (e.g., gold_slot_performance):

📖 Overview Tab¶

• Asset description
• Owner and steward
• Classifications applied
• Glossary terms linked

Schema Tab¶

• Column names and data types
• Column-level classifications
• Column descriptions

Lineage Tab¶

• Upstream data sources
• Downstream consumers
• Transformation steps

2.3 Asset Hierarchy View¶

Fabric-Casino-POC (Source)
├── lh_bronze (Lakehouse)
│   ├── bronze_slot_telemetry (Table)
│   ├── bronze_player_profile (Table)
│   ├── bronze_financial_txn (Table)
│   └── ...
├── lh_silver (Lakehouse)
│   ├── silver_slot_cleansed (Table)
│   ├── silver_player_master (Table)
│   └── ...
├── lh_gold (Lakehouse)
│   ├── gold_slot_performance (Table)
│   ├── gold_player_360 (Table)
│   └── gold_compliance_reporting (Table)
├── Casino Analytics Model (Semantic Model)
└── Casino Executive Dashboard (Report)

🛠️ Step 3: Apply Classifications¶

3.1 Built-in Classifications¶

Purview includes 200+ built-in classifications for:

3.2 Review Auto-Classifications¶

1. Navigate to table: silver_player_master
2. Click Schema tab
3. Review automatically detected classifications:

Tip: Auto-classification uses pattern matching and ML. Review for accuracy in your specific context.

3.3 Create Custom Classifications¶

For casino-specific data patterns:

1. Go to Data Map > Classifications
2. Click + New
3. Create custom classifications:

Classification: Casino Player ID¶

Classification: Gaming Transaction¶

Classification: Regulatory Filing¶

3.4 Apply Classifications to Assets¶

1. Open table silver_player_master
2. Click Edit
3. Under Classifications, add:
4. Casino - Player PII
5. Confidential
6. Click Save

For compliance tables: 1. Open gold_compliance_reporting 2. Add classifications: - Casino - Regulatory Filing - Highly Confidential - Regulatory - BSA/AML

🛠️ Step 4: Create Business Glossary¶

4.1 Navigate to Glossary¶

1. Go to Data Catalog > Glossary
2. Create a hierarchical structure for casino terms

4.2 Create Term Hierarchy¶

Casino Glossary (Root)
├── Gaming Operations
│   ├── Slot Metrics
│   │   ├── Coin In
│   │   ├── Coin Out
│   │   ├── Net Win
│   │   └── Hold Percentage
│   └── Table Games
│       ├── Drop
│       └── Handle
├── Player Management
│   ├── Player Value
│   │   ├── Theoretical Win
│   │   └── ADT (Average Daily Theo)
│   └── Loyalty Tiers
├── Compliance & Regulatory
│   ├── CTR (Currency Transaction Report)
│   ├── SAR (Suspicious Activity Report)
│   └── W-2G (Gambling Winnings)
└── Financial
    └── Cash Handling

4.3 Create Business Terms¶

Term: Coin In¶

Name: Coin In
Definition: |
  Total amount wagered by players on a gaming device.
  Represents the handle or total bets placed, regardless of outcome.
  This is the primary volume metric for slot operations.

Acronym: CI
Status: Approved
Owner: Slot Operations Manager

Related Terms:
  - Coin Out
  - Net Win
  - Hold Percentage

Formula: Sum of all wagers placed

Regulatory Reference: NIGC MICS Section 543.24

Resources:
  - Gaming Commission Handbook
  - Internal Slot Accounting Manual

Term: Theoretical Win (Theo)¶

Name: Theoretical Win
Definition: |
  The statistical expected win from a player or device based on
  the mathematical house advantage and total wagers.
  Used for player rating and marketing decisions.

Synonyms:
  - Theoretical
  - Expected Win
  - Theo

Formula: Theo = Coin In x House Edge

Example: |
  If player wagers $10,000 on a slot with 8% hold:
  Theo = $10,000 x 0.08 = $800

Status: Approved
Owner: Player Development

Related Terms:
  - Hold Percentage
  - Actual Win
  - Player Value Score

Term: Currency Transaction Report (CTR)¶

Name: Currency Transaction Report
Definition: |
  Federal filing required for cash transactions of $10,000 or more
  in a single gaming day. Must be filed within 15 days of transaction.
  Casinos must aggregate multiple transactions by same person.

Acronym: CTR
Status: Approved
Owner: Compliance Officer

Regulatory Reference: 31 CFR 1021.311

Filing Requirements:
  - Amount: $10,000 or more in cash
  - Timeframe: Single gaming day
  - Deadline: 15 calendar days
  - Form: FinCEN Form 112

Related Terms:
  - SAR
  - BSA
  - Gaming Day
  - AML

Term: Hold Percentage¶

Name: Hold Percentage
Definition: |
  The percentage of total wagers retained by the casino.
  Calculated as Net Win divided by Coin In.
  Primary profitability metric for slot operations.

Synonyms:
  - Hold
  - Win Percentage
  - House Edge (theoretical)

Formula: Hold % = (Coin In - Coin Out) / Coin In x 100

Typical Ranges:
  - Slots: 6-12%
  - Video Poker: 2-6%
  - Table Games: 2-5%

Status: Approved
Owner: Finance

Monitoring: |
  Significant deviations from expected hold may indicate:
  - Machine malfunction
  - Cheating/fraud
  - Statistical variance (short term)

4.4 Assign Terms to Data Assets¶

1. Open table gold_slot_performance
2. Click Edit
3. Map glossary terms to columns:

1. Click Save

Best Practice: Assign glossary terms at both table and column level for complete documentation.

🛠️ Step 5: View Data Lineage¶

5.1 Navigate to Lineage¶

1. Open any Gold table (e.g., gold_slot_performance)
2. Click Lineage tab
3. View the data flow visualization

5.2 Understanding Lineage View¶

flowchart LR
    subgraph Sources["External Sources"]
        SMS[Slot Management System]
        CMS[Casino Management System]
    end

    subgraph Bronze["Bronze Layer"]
        BST[bronze_slot_telemetry]
        BPP[bronze_player_profile]
        BFT[bronze_financial_txn]
    end

    subgraph Silver["Silver Layer"]
        SSC[silver_slot_cleansed]
        SPM[silver_player_master]
    end

    subgraph Gold["Gold Layer"]
        GSP[gold_slot_performance]
        GP3[gold_player_360]
        GCR[gold_compliance_reporting]
    end

    subgraph Consumption["Consumption"]
        SM[Casino Analytics Model]
        RPT[Executive Dashboard]
    end

    SMS --> BST --> SSC --> GSP --> SM --> RPT
    CMS --> BPP --> SPM --> GP3 --> SM
    BFT --> GCR

5.3 Complete Lineage for Slot Performance¶

View the full transformation chain:

Slot Management System (External)
        │
        ▼
bronze_slot_telemetry
  • Raw event data
  • Added: _bronze_ingested_at, _bronze_batch_id
        │
        ▼
silver_slot_cleansed
  • Data quality validation
  • Null handling
  • Type standardization
  • Added: _silver_processed_at
        │
        ▼
gold_slot_performance
  • Daily aggregation by machine
  • Calculated metrics (hold %, theo)
  • Added: _gold_computed_at
        │
        ▼
Casino Analytics Model (Semantic Model)
  • DAX measures
  • Relationships
        │
        ▼
Casino Executive Dashboard (Report)
  • Visualizations
  • KPIs

5.4 Impact Analysis¶

Use lineage for change impact assessment:

1. On silver_slot_cleansed, click View Lineage
2. Right-click on a downstream asset
3. Click Impact Analysis
4. View all dependent assets

Example Impact Report:

If silver_slot_cleansed schema changes: - gold_slot_performance (IMPACTED) - Casino Analytics Model (IMPACTED) - Executive Dashboard (IMPACTED) - 3 additional reports (IMPACTED)

Warning: Always run impact analysis before making schema changes to understand downstream effects.

🛠️ Step 6: Sensitivity Labels and Data Policies¶

6.1 Apply Sensitivity Labels¶

1. In Purview, go to Information Protection
2. Configure labels for Fabric assets:

1. Apply to specific tables:

6.2 Configure Data Access Policies¶

1. Go to Data Policy > Data access policies
2. Create policy for analysts:

1. Create policy for compliance team:

6.3 PII Handling and Masking¶

For player PII, implement data masking:

# Example: Mask PII in Silver layer
from pyspark.sql.functions import sha2, concat, lit, regexp_replace

def mask_pii(df):
    """Apply PII masking for non-authorized users"""
    return df \
        .withColumn("ssn_hash", sha2(col("ssn"), 256)) \
        .withColumn("email_masked",
            regexp_replace(col("email"), "^(.{2}).*@", "$1***@")) \
        .withColumn("phone_masked",
            concat(lit("***-***-"), substring(col("phone"), -4, 4))) \
        .drop("ssn", "email", "phone")

🛠️ Step 7: Monitoring and Compliance Reporting¶

7.1 Data Estate Insights¶

1. Go to Data Estate Insights
2. Review key dashboards:

7.2 Generate Compliance Report¶

For regulatory audits:

1. Go to Data Catalog > Browse
2. Apply filters:
3. Classification = Casino - Regulatory Filing
4. Classification = Highly Confidential
5. Export asset list
6. Include in compliance documentation

Sample Compliance Report Structure:

# Casino Data Governance Compliance Report
## 📅 Period: Q4 2024

### 1. Data Inventory
- Total cataloged assets: 45
- Tables with PII: 12
- Compliance-related tables: 8

### 2. Classification Coverage
- Assets with sensitivity labels: 100%
- Assets with glossary terms: 85%
- Assets with assigned owners: 100%

### 3. Access Controls
- Data access policies: 5
- Users with compliance data access: 8
- Access reviews completed: Quarterly

### 4. Lineage Documentation
- End-to-end lineage mapped: Yes
- Source systems documented: Yes
- Transformation logic documented: Yes

### 5. Audit Trail
- Scan frequency: Weekly
- Last successful scan: [Date]
- Classification rule updates: [List]

7.3 Audit Trail Export¶

1. Go to Monitoring > Diagnostics
2. Review and export:
3. Scan history
4. Classification changes
5. Policy changes
6. Access requests

# Using Purview REST API for audit export
import requests

endpoint = "https://your-purview.purview.azure.com"
headers = {"Authorization": f"Bearer {token}"}

# Get all assets with regulatory classification
response = requests.post(
    f"{endpoint}/catalog/api/atlas/v2/search/advanced",
    headers=headers,
    json={
        "keywords": "*",
        "filter": {
            "and": [
                {"classification": "Casino - Regulatory Filing"},
                {"entityType": "azure_datalake_gen2_path"}
            ]
        },
        "limit": 100
    }
)

regulatory_assets = response.json()

✅ Validation Checklist¶

Before moving to the next tutorial, verify:

• Purview Catalog Populated - All Fabric assets scanned and cataloged
• Lineage Visible - End-to-end data flow mapped from Bronze to Reports
• Classifications Applied - Auto and custom classifications on sensitive data
• Glossary Terms Created - Casino business terminology defined (min 10 terms)
• Access Policies Active - Data access policies configured and enforced
• Compliance Report Generated - Regulatory assets documented