Self service analytics and insights (Preview)

Self-service analytics and insights refer to data, tools, and platforms that enable business users to access, analyze, and generate insights from data independently. The Purview data governance application publishes domain and dimension models of metadata and metadata models into OneLake, allowing customers to analyze and generate insights by bringing their own tools and compute. Self-service analytics of data governance metadata is valuable in driving continuous improvement of customers' data estate health management and fostering a data-driven culture across the organization by democratizing access to data estate insights and health management.

Key Components

• Data Models: 3NF Models with domains and dimensions details
• Metadata: Data governance metadata that includes:
  • Governance domains
  • data products
  • data assets
  • glossary terms
  • subscription request
  • data quality rules
  • dimensions
  • data quality scores and more are coming soon.

Benefits

• Empowerment: Empowers data practitioners, data product owners, data stewards, and analysts to explore data governance metadata and link metadata from various sources to derive insights.
• Flexibility and Efficiency: Customer will able to create custom reports in addition to out-of-the-box reports in health management.
• Agility: Allows customers organizations to respond more swiftly to health management issues and remediation.
• Cost-Effective: Reduces the need for setting up platforms and build tools. All data is available in OneLake and customer will able to use available tools (Fabric semantic model, PBI reporting, Data flow, and notebook) in OneLake.

Currently available reports (out-of-the-box)

Here are the available out-of-the-box reports. These reports aren't customizable.

• Classic assets - an overview of assets by type and collection, and their curation status.
• Classic catalog adoption - to understand at a glance how your data catalog is being used. your glossary, providing a snapshot of terms and their status.
• Classic classifications - an overview of assets classified and the types of classifications.
• Classic data stewardship - an overview of assets classified and the types of classifications.
• Classic glossary - health and use of glossary terms.
• Classic sensitivity labels -an overview of assets that have sensitivity labels applied and the types of labels applied.
• Data governance - The data governance health report allows your team to track your health progress at a glance, and identify areas that need more work.
• Data quality health - Data quality dimensions and data quality rules performance reporting.

Subscribe Microsoft Purview Catalog metadata

You can subscribe to Microsoft Purview's data governance metadata for analytics and to derive insights. To subscribe, follow the below listed steps:

1. Select Settings from the left pane, select Data catalog, then select solution integrations.

   

2. Select Edit.

3. Add Storage type.

4. Add Location URL

5. Add folder name at the end of the URL - example: /DEH (see the screenshot)

6. Grant contributor accesses to Microsoft Purview Manage Service Identity (MSI) to your fabric workspace.

7. Test connection.

   

Data model of self-serve analytics metadata

The 3NF domain model is part of the normalization process in relational database design, which ensures that the database is free of redundancy and update anomalies. A database schema is in the Third Normal Form if it meets the requirements of the First and Second Normal Forms and all its attributes are functionally dependent only on the primary key. The purpose of 3NF domain model is used to structure data in a way that minimizes duplication and ensures data integrity. It focuses on breaking down data into smaller, related tables where each piece of information is stored only once.

Characteristics:

• Elimination of Transitive Dependencies: Nonkey attributes shouldn't depend on other nonkey attributes.
• Logical Grouping: Data is logically grouped into domains based on its function or meaning.
• Entity-Relationship Diagrams (ERDs): Commonly used to represent 3NF domain models, showing how entities relate to each other.

Create a semantic model

A semantic model in the context of data and analytics refers to a structured representation of data that defines the meaning, relationships, and rules within a specific domain. It provides a layer of abstraction that helps users understand and interact with complex data by making it more intuitive and accessible, especially in the context of business intelligence (BI) and analytics platforms. A semantic model is always required before any reports can be built. Within the warehouse, a user can add warehouse objects - tables or views to their default Power BI semantic model. They can also add other semantic modeling properties, such as hierarchies and descriptions. These properties are then used to create the Power BI semantic model's tables. Users can also remove objects from the default Power BI semantic model.

To reate semantic model from the Microsoft Purview Data Governance metadata domain and dimension models:

1. Open a warehouse in your Fabric workspace.

2. Publish the domains and dimensions models files to delta tables

   1. Select the ellipsis button (...)
   2. Select **Load to Tables > New table
   3. New table name prepopulated
   4. Select File type parquet

   

3. You can also use shortcut to shortcut the domain and dimensions model from OneLake and within OneLake

   1. Select the ellipsis button (...) of Tables
   2. Select New Shortcut and select Microsoft OneLake from New Shortcut sources page.
   3. Select domains and dimensions models table to shortcut

   

After you've published all files to delta tables either manually or via shortcut, you're able to add the delta tables to semantic model.

1. Switch to SQL analytics endpoint page from the Lakehouse page.
2. Select Reporting from top left corner of the SQL analytics endpoint page.
3. Select Manage default semantic model.
4. Select the tables from dbo > Tables that you want to add to the semantic model for reporting.

To add objects such as tables or views to the default Power BI semantic model, select Automatically update semantic model.

Create a Power BI report

Power BI is natively integrated in the whole Fabric experience. This native integration includes a unique mode, called DirectLake, for accessing the data from the lakehouse to provide the most performant query and reporting experience. DirectLake is a groundbreaking new capability that allows you to analyze very large semantic models in Power BI. With DirectLake, you load parquet-formatted files directly from a data lake without needing to query a data warehouse or lakehouse endpoint, and without needing to import or duplicate data into a Power BI semantic model. DirectLake is a fast path to load the data from the data lake straight into the Power BI engine, ready for analysis.

In traditional DirectQuery mode, the Power BI engine directly queries the data from the source for each query execution, and the query performance depends on the data retrieval speed. DirectQuery eliminates the need to copy data, ensuring that any changes in the source are immediately reflected in query results.

For more details follow the guideline: how to create a Power BI report in Microsoft Fabric.

Coming soon

• ADLSg2 support is currently in private preview, public preview is planned for November 30, 2024.
• Scheduling the data refreshing job is not yet supported. Tentative ETA for this feature is 30th November.

Related content

• OneLake shortcuts
• Fabric Data Activator
• What is data activator
• Configure Fabric Copilot
• Build automation with PowerAPP