Data Governance and Discovery with Microsoft Purview
Data governance, Data Map, scans, metadata, classification, lineage, quality, Unified Catalog, governance domains, data products, and responsible use
Suggested study time: 23 minutes • Beginner level • Aligned with the SC-900 study guide and official Microsoft Learn documentation
By João Ricardo Dutra••Complete material
1. Introduction
From the administration of banks to the management of data assets
Data governance did not emerge as an isolated product. Its roots are in data dictionaries, bank administration, record management, and the early quality programs and master data management. With the expansion of the internet, big data, cloud computing, and artificial intelligence, data began to circulate across dozens of platforms, teams, and countries. Knowing only where a file was stored became insufficient: it became necessary to understand the meaning, origin, quality, responsible party, and legitimate conditions of use of each data set.
For the reader, this knowledge transforms an apparently chaotic environment into an understandable system. For society, well-governed data increases the reliability of financial services, health, education, research, and public policies, as well as reducing risks of discrimination, leaks, and decisions based on incorrect information. In an era in which AI models learn from large volumes of data, the quality of the input directly influences confidence in the output.
Throughout this chapter, the abstract idea of "data governance" will be converted into concrete elements: assets, metadata, classifications, owners, lineage, data products, governance domains, and quality rules. The guiding question for the reader is simple: how does a person find the right data, understand its history, and know if they can use it safely?
Central idea
Invisible data, without an owner and without context, is not a reliable asset. Governance creates visibility, accountability, and usage criteria.
2. What is data governance
Data governance is the system of decisions, responsibilities, policies, standards, and controls that guides how data is created, described, protected, shared, used, maintained, and disposed of. It defines who decides, who executes, how quality is measured, what uses are acceptable, and how conflicts are resolved.
Good governance does not try to centralize all decisions in a single team. The modern model usually combines corporate standards with distributed responsibility across business domains. This federated approach allows Finance, Customers, Risk, or Human Resources to maintain context and ownership of their data, while a central office establishes common principles, roles, and metrics.
Concept
Main question
Example
Data governance
Who decides and according to which rules?
Define owner, steward, access policy, and quality criteria.
Data management
How is data operated during the life cycle?
Modeling, integration, backup, storage, and archiving.
Data security
How to prevent unauthorized access, alteration, or loss?
Encryption, RBAC, DLP, monitoring, and response.
Compliance
Which legal, regulatory, and contractual obligations need to be met?
LGPD, retention, audit and evidence.
Data catalog
How to locate and understand assets and metadata?
Search for tables, reports, owners, descriptions, and lineage.
Do not confuse Governance defines the model of responsibility and decision-making. A catalog tool supports this model, but does not replace people, policies, and processes.
3. Fundamentals of governance
Figure 1 - Cycle that transforms scattered data into reliable and usable information.
3.1 Discovery and cataloging
Discovery is the ability to locate sources and assets in a data estate. Cataloging is the systematic organization of the metadata of these assets so that they can be searched, understood, and governed. A catalog does not need to copy the content of all databases; it gathers descriptions, schemas, classifications, relationships, and accountability information.
3.2 Classification and context
Classifying is associating categories with data based on technical patterns, semantics, or business rules. A column can be classified as an email address, tax identifier, or account number. The business context explains why this data exists, which process produces it, and which policies should accompany it.
3.3 Ownership, quality, and use
Ownership establishes accountability. Quality measures whether the data is suitable for the purpose. Responsible use combines legitimate purpose, minimization, security, transparency, and respect for applicable policies. These elements reinforce each other: without an owner, quality issues remain unresolved; without context, users may interpret the data incorrectly.
4. Roles and responsibilities
Figure 2 - Complementary roles in a data governance program.
The data owner is usually a business authority responsible for the value, risk, quality, and access rules of a domain or product. The data steward translates decisions into daily practice: maintains definitions, resolves inconsistencies, monitors quality, and helps users interpret the assets. The data custodian operates the technical controls of the platform, but should not decide alone the business purpose of the data.
Consumers also have responsibility. Finding a dataset in the catalog does not mean that all use is authorized. The user must respect the purpose, terms, classification, privacy requirements, and access conditions. The central data office coordinates the model, sets standards, and measures maturity without taking away the specialized knowledge from the domains.
Paper
Typical responsibility
Risk if absent
Data Office
Standards, operational model, metrics, and coordination.
Fragmented governance and incompatible criteria.
data owner
Decision about value, risk, quality, and access.
Assets without accountability or priority.
data steward
Curation, glossary, quality, and consistency.
Outdated catalog and ambiguous concepts.
Custodian
Technical operation, availability, and protection.
Operational failures, excessive access, or data loss.
Consumer
Use according to purpose and terms.
Incorrect interpretation or misuse.
5. Metadata: data about data
Figure 3 - Metadata layers that help describe a data asset.
Metadata is information that describes other data. A table name, the type of a column, the owner, the time of an execution, and the relationship between a source and a report are examples. The value of a catalog depends on the richness, timeliness, and consistency of this metadata.
Technical metadata helps engineers understand structures. Business metadata allows analysts to search using their own language. Operational metadata shows processes and executions. Semantic metadata connects assets to classifications, terms, and relationships. The Microsoft Purview Data Map brings these dimensions together in a metadata graph.
Active metadata
In the modern model, metadata is not only used for documentation. It can guide search, access, quality, automation, and governance policies.
6. Governance architecture in Microsoft Purview
The current governance experience of Microsoft Purview has two central solutions. The Data Map is the technical foundation for capturing, storing, and relating metadata. The Unified Catalog is the SaaS experience aimed at consumption, curation, quality, health, and business value.
Solution
Main function
Audience and activities
Microsoft Purview Data Map
Build the technical map and the metadata graph.
Administrators and curators record sources, configure scans, organize domains or collections, and monitor ingestion.
Microsoft Purview Unified Catalog
Transform metadata into a governance and discovery experience for the business.
Owners, stewards, and consumers work with domains, products, glossary, quality, health, search, and access.
The two solutions are complementary. Without Data Map, the catalog does not have a comprehensive technical view of the assets. Without Unified Catalog, the technical inventory may remain difficult to interpret and use for business users. The expected flow is to capture metadata, curate it, add context, and make reliable products available.
Summary for the SC-900
Data Map = technical foundation of metadata. Unified Catalog = business-oriented discovery, curation, and governance experience.
7. Microsoft Purview Data Map
Figure 4 - Conceptual components of the Data Map.
Microsoft Purview Data Map captures metadata from analytical, operational, and SaaS systems in Azure, on-premises, hybrid, and multicloud environments. It stores assets and relationships in a graph structure, allowing representation of owners, stewards, hierarchies, classifications, and lineage.
An asset can be a table, file, database, report, model, or other object recognized by the connector. The map can be enriched automatically through scans and integrations and manually through curation. The capacity is expressed by metadata storage and operation throughput, with elasticity according to consumption.
7.1 Domains and collections in the Data Map
In Data Map, domains and collections help organize sources, scans, assets, and administrative responsibilities. Collections form hierarchies and can act as access boundaries for metadata. These technical objects should not be confused with governance domains in the Unified Catalog, which are context and business ownership boundaries for products and concepts.
8. Recording, scans, and ingestion
Figure 5 - Flow of registration, scanning, ingestion, and curation.
8.1 Registering is not scanning
Registering a source informs Microsoft Purview where it exists and in which domain or collection it will be organized. Registration alone does not extract the schema. The scan connects to the source using a supported authentication method, traverses the configured scope, and captures metadata.
8.2 Scan Levels
The current documentation describes scan levels. A basic level can capture name, size, and identifier; an intermediate level extracts schemas when available; a more complete level also evaluates samples against classification rules. The effective level depends on the source and configuration.
8.3 Ingestion
After the scan, the ingestion processes the metadata and loads it into the Data Map. It can also receive lineage from connected services, such as integration platforms. A completed scan does not necessarily mean that all assets are already available in the catalog: the ingestion needs to finish.
Scan security
Credentials, integration runtime, connectivity, and permissions must follow the principle of least privilege. Catalog governance does not justify unrestricted access to sources.
9. Classification, cataloging, and curation
Classifications help to recognize patterns in metadata and, depending on the source and scan level, in sampled values. Examples include email, phone, identifiers, and financial data. They are useful for discovery and prioritization, but need to be assessed in context: a technical match may not, by itself, represent the actual purpose or risk of the asset.
Element
What does it represent
Example
Classification
Technical or semantic category applied to an asset or column.
Email address, card, personal identifier.
Description
Human explanation about content and purpose.
Consolidated table of active clients.
Owner or expert
Contact responsible for the asset or for its understanding.
Customer Data Team.
Glossary term
Standardized business definition.
Active customer, net revenue, default.
Sensitivity label
Protection and sensitivity brand integrated into the Purview ecosystem.
Confidential - Personal Data.
Curation is the work of improving the meaning and usefulness of metadata. a catalog may have thousands of assets discovered automatically, but it will remain of little use if technical names do not have description, owner, or connection to business concepts. The steward transforms inventory into organizational knowledge.
Trap of evidence
Classification, glossary term, and sensitivity label are related, but they are not synonyms.
10. Data lineage
Figure 6 - Lineage from origin to consumption.
Lineage describes how data arises, moves, and is transformed until it reaches its destination. It can show relationships at the asset level and, in compatible integrations, at the column level. Lineage can be captured through scans, native pipeline connections, or integration APIs.
10.1 Impact Analysis
Before changing a column in the source, the team can check reports, models, and dependent products. This reduces changes that break downstream consumers. Impact analysis answers 'what will be affected if I change this asset?'.
10.2 Root Cause Analysis
When an indicator shows an incorrect value, the team can navigate backward, identify transformations, and locate the origin point of the problem. The root cause answers 'where did this error come from?'.
10.3 Audit and trust
Lineage helps to explain the provenance of a number, to demonstrate controls, and to assess whether an AI output or report uses approved sources. It does not replace detailed operational logs, but it connects metadata to provide an understandable view of the data's journey.
11. Ownership and stewardship
Property should not just be a field filled in to increase the completeness of the catalog. An owner needs to have the authority to make decisions about access, correction priority, quality criteria, and lifecycle. A steward needs time, processes, and metrics to keep the asset reliable.
Decision
Owner
Steward
Custodian
Purpose and business value
Approves and responds.
Documents and guides.
Implements technical support.
Quality criterion
Define acceptable level.
Sets up and monitors rules.
Operates corrections and pipelines.
Access
Approves policy or criteria.
Validate context and terms.
Applies technical controls.
Data incident
Decide impact and priority.
Investigates metadata and usage.
Contains and corrects platform.
Obsolescence
Approves discontinuation.
Updates catalog and consumers.
Archives or technically removes.
Federated governance distributes these roles by domains, but maintains corporate standards. The goal is not to create bureaucracy, but rather to eliminate the common state in which everyone uses a piece of data, but no one is responsible for its definition or quality.
Accountability
A critical asset without an owner is an organizational risk, even if it is technically protected.
12. Data quality
Figure 7 - Cycle of quality definition, evaluation, and improvement.
Quality is the suitability of the data for the declared purpose. A record may be sufficient for sending a communication, but inadequate for regulatory calculation. Therefore, metrics need to be linked to the use and risk of the data product.
Dimension
Question
Example of rule
Completeness
Are the required fields filled in?
Percentage of non-null CPF greater than 99%.
Validity
Do the values follow the format or domain?
Valid date of birth and not in the future.
Uniqueness
Are there undue duplications?
A customer identifier per person.
Consistency
Do related sources agree?
Client status is the same in the CRM and in the data warehouse.
Accuracy
Does the value correctly represent reality?
Address confirmed by a reliable source.
Current events
Is the data recent enough?
Load completed in the last 24 hours.
In the Unified Catalog, quality rules can produce scores at the asset, product, and governance domain levels. Problems and actions provide visibility into what needs to be corrected. The score helps guide improvement, but it does not eliminate the need for human judgment and validation of the process that generates the data.
13. Microsoft Purview Unified Catalog
Figure 8 - Resources that connect business context, discovery, and responsible use.
The Microsoft Purview Unified Catalog is the governance and discovery experience built on the metadata inventory. It organizes data by business context, allows grouping assets into products, standardizing vocabulary, tracking quality and health, searching information, and providing access workflows.
The proposal is to serve consumers, owners, and stewards in an integrated experience. Instead of navigating through a flat list of tables, the user can explore a governance domain, find a product associated with a purpose, and evaluate its description, owner, terms, assets, lineage, and quality.
Current context
The new experience of the Unified Catalog is being rolled out gradually and depends on the enterprise version and regional availability. Features and preview statuses may change.
14. Governance domains
A governance domain is a business boundary that organizes ownership, discovery, and application of governance practices. It can represent Finance, Marketing, Customers, a product, a corporate entity, a regulatory obligation, or a project. It functions as a mini catalog oriented to the context of that area.
The domain contains owners and concepts such as data products, glossary terms, OKRs, and critical data elements. The idea is to bring governance closer to the teams that understand the business, without losing corporate standards. This reduces the bottleneck caused by a central team that would need to curate every asset across the company.
Governance domain
Data Map domain/collection
Organizes products and concepts from a business perspective.
Organizes sources, scans, assets, and administrative access to metadata.
There are owners and business stewards.
It has Data Map functions and hierarchical administration limits.
Supports discovery, policies, and product value.
Supports technical operation, delegation, and isolation of map resources.
Example: Clients or Credit Risk.
Example: Brazil unit, Production environment, or CRM collection.
The essential distinction of the Governance domain of the Unified Catalog is not just a technical folder. It represents responsibility, language, and business value.
15. Data products
A data product is a logical grouping of assets related to a use case. It can bring together tables, files, reports, models, and documentation needed for a purpose, such as '360 Customer View' or 'Monthly Regulatory Indicators.' The product adds context and reduces the work of searching for each component separately.
Product component
Purpose
Name and description
Explain what the product delivers and what problem it exists for.
Business use
Define purpose, audience, and supported decisions.
Owner and contacts
Provide accountability and support.
Associated assets
Gather necessary sources, tables, files, and reports.
Glossary and CDEs
Standardize meaning and highlight critical elements.
Quality and health
Demonstrate confidence and pending actions.
Terms and access
Inform conditions of use and facilitate access requests.
A data product is not necessarily a copy of the data nor a new database. It is a governed and value-oriented packaging, which points to existing assets and provides everything the consumer needs to assess and use these assets responsibly.
Data product x asset
An asset is an individual object. A data product combines assets and context for a reusable purpose.
16. Glossary, CDEs and objectives
16.1 Glossary terms
Glossary terms create a common vocabulary. They translate technical names, reduce ambiguities, and allow different areas to agree on concepts such as “active customer,” “recognized revenue,” or “critical incident.” In the current experience, terms can be active objects that also carry policies and governance guidelines.
16.2 critical data elements
critical data elements, or CDEs, represent important elements that may appear with different names in different systems. A "Customer ID" concept can relate to the columns CustID, ClientNumber, and CID. This abstraction helps to standardize, apply quality rules, and handle critical data consistently.
16.3 OKRs
Objectives and Key Results connect governance to value. Instead of measuring only the number of cataloged assets, the organization can track objectives such as reducing the time to find reliable data or increasing the quality of products used by AI models.
Object
Question that answers
Glossary term
What does this concept mean for the business?
Critical Data Element
Which technical fields represent this critical element?
OKR
What business outcome should governance produce?
Classification
What pattern or category was identified in the asset?
17. Discovery, search, and access
Data discovery is the ability of a person to locate relevant information without previously knowing the server, database, or technical name of the table. In the Unified Catalog, the search can consider name, description, governance domain, glossary, critical data elements, owner, and other attributes. Current features also include natural language search, as available.
Finding does not mean automatically accessing. The catalog can expose metadata for discovery while keeping the content protected at the source. Access requests and policies help balance self-service with security, purpose, and right-use. Actual controls continue to depend on the data platforms and the available integrations.
Stage
Consumer question
Useful information in the catalog
Search
Is there a product for my problem?
Name, description, domain, glossary, and use case.
Evaluate
Can I trust and interpret correctly?
Owner, quality, lineage, timeliness, and terms.
Request access
Who approves and what conditions apply?
Policy, purpose, contact, and request flow.
Consume
How to use without breaking rules?
Terms of use, classification, and context.
Give feedback
How to report an error or need?
Owner, steward, and support channels.
Principle of least privilege The goal of self-service is not to remove controls, but to reduce friction for approved and traceable uses.
18. Heritage health and responsible use
Data health expands the view beyond a single quality rule. Health controls measure governance practices, scores show progress, and actions indicate necessary corrections. A healthy asset has understandable assets, defined owners, complete products, executed rules, proper access, and updated metadata.
Responsible use means using data for a legitimate, proportional, and transparent purpose. It includes respecting privacy, avoiding excessive collection, assessing biases, maintaining security, documenting limitations, and not reusing data outside the approved context. In the AI era, governance is part of the model's own security: incorrect data, without consent or provenance, can produce harmful results.
Sign of health
Governance question
Owner defined
Is there someone responsible for decisions and corrections?
Description and glossary
Does a user understand meaning and purpose?
Available lineage
Is it possible to explain origin and dependencies?
Measured quality
Are there rules and outcomes appropriate for the use?
Governed access
Does the consumption occur by authorized people and purposes?
Followed actions
Do problems have a person in charge, priority, and deadline?
AI Governance A sophisticated model does not make up for data without quality, context, permission, or representativeness.
19. Integrated practical scenario
Figure 9 - Governance flow of the 360 Customer Vision product.
A financial institution keeps customer data in the CRM, transactions in Azure SQL, histories in the data lake, and reports in Power BI. Analysts spend days searching for sources and do not know which table is official. The first step is to register and scan the sources in the Data Map. The scans capture schemas, classifications, and assets; pipeline integrations enrich the lineage.
In the Unified Catalog, the organization creates the governance domain "Clients", assigns an owner and stewards, and defines terms such as Active Client and Client Identifier. Relevant assets are grouped in the data product "360° Client View". Completeness, uniqueness, and timeliness rules generate scores. The description informs purpose and limitations; the access policy requires justification and approval.
A data scientist searches for customer data in business language, finds the product, examines quality and lineage, and requests access for a retention model. The owner evaluates the purpose, access is granted according to policy, and consumption is associated with a governed set. The gain is not just technological: the organization reduces rework, improves traceability, and makes responsibility for the data explicit.
20. Implementation and best practices
Start with use cases and priority domains, not with an attempt to catalog everything without purpose.
Define owners and stewards before requiring mass metadata completion.
Automate discovery and classification, but maintain human curation for context and quality.
Associate quality rules with the purpose and risk of each product.
Integrate cataloging into pipelines, architecture, security, privacy, and change processes.
Measure business results, such as discovery time, reuse, incident reduction, and trust.
Review metadata, products, and policies to avoid a catalog that is technically complete but outdated.
20.1 Frequent errors
Error
Consequence
Correction
Catalog without use case
Large volume of assets without consumers.
Prioritize products and discovery journeys.
Confuse owner with administrator
Business decisions remain with those who only operate the platform.
Separate accountability from technical custody.
Treat quality as a unique project
Scores degrade after the initial delivery.
Create continuous monitoring and actions.
Expose metadata without governance
Sensitive information or critical relationships become excessively visible.
Apply roles, domains, collections, and least privilege.
Ignore lineage
Changes break reports and causes remain hidden.
Integrate pipelines and review dependencies.
Use glossary without adoption
Terms exist, but they do not change the language of business.
Involve specialists and apply terms to products.
21. Important comparisons for the SC-900
Concepts
Essential difference
Data Map vs Unified Catalog
Data Map captures and relates metadata; Unified Catalog offers business-oriented discovery and governance.
Governance domain x collection
Governance domain organizes business context and responsibility; collection organizes metadata and administrative access in the Data Map.
Asset x data product
An asset is an individual object; a data product combines assets for a purpose.
Classification x glossary term
Classification identifies category; glossary term defines business meaning.
Lineage x audit log
Lineage represents the flow and transformation of data; audit log records activities of users and administrators.
Owner vs. custodian
Owner is responsible for value and decisions; custodian operates technical controls.
Quality score x guarantee
Score guides confidence and improvement, but it does not prove that the data is perfect for every use.
Discovery x access
Finding metadata does not automatically grant access to the content.
Technical map memorization -> Data Map. Business catalog -> Unified Catalog. Grouping by use -> data product. Organizational context -> governance domain. Data history -> lineage.
22. Quick review
Term
Objective memorization
Data governance
Decisions, roles, standards, and controls to create value and reduce risk.
Data Map
Distributed asset metadata graph.
Scan
Captures metadata, schema, and classifications, according to support.
Ingestion
Processes and loads metadata into the Data Map.
Metadata
Information that describes data and its relationships.
Lineage
Origin, movement, transformation, and destination.
Unified Catalog
Business-oriented SaaS experience in discovery and governance.
Governance domain
Context limit, property, and governance.
Data product
Asset package for a use case.
Glossary term
Standardized vocabulary and business context.
CDE
Logical representation of a critical element in different systems.
Data quality
Measurable suitability for the purpose.
Data health
Maturity view, controls, and actions of the assets.
Responsible use
Legitimate, safe, transparent use and in accordance with purpose.
Final mind map Discover -> Data Map. Understand -> metadata, classification, and lineage. Organize -> governance domains and data products. Trust -> owner, glossary, and quality. Use -> search, access, and responsible use.
23. Conclusion
Data governance connects technology, responsibility, and value. Discovery reveals the assets; cataloging organizes metadata; classification identifies categories; ownership establishes accountability; quality measures fitness; lineage explains the journey; and responsible use defines limits for consumption. When these elements are treated separately, gaps remain. When they work as a system, data becomes more reliable and reusable.
The Microsoft Purview Data Map provides the technical foundation by capturing metadata and relationships in on-premises, hybrid, and multicloud environments. The Unified Catalog uses this foundation to create a business-oriented experience with governance domains, data products, glossaries, search, quality, health, and access. The main distinction for the exam is to understand that the map describes the assets, while the catalog transforms this description into a governance and discovery experience.
In my assessment, the greatest benefit of modern governance is not producing documentation: it is reducing the distance between those who create, those who protect, those who understand, and those who use the data. In a society increasingly dependent on automated decisions, reliable and responsible data ceases to be an operational detail and becomes a condition for sustainable innovation.
End of the SC-900 trail
This chapter concludes the 17 chapters of the track. The final review should connect identity, security, compliance, and governance as complementary responsibilities in the Microsoft cloud.
24. Review questions
Question 1: Which statement correctly describes the Microsoft Purview Data Map?
A) It is a repository that replaces all databases. B) It is the foundation that captures and relates metadata of distributed assets. C) It is an exclusive email protection tool. D) It is just an audit dashboard.
Commented answer
Correct answer: B. The Data Map maintains a graph of metadata, classifications, relationships, and lineage over the data assets.
Question 2: What is the main purpose of a data product in the Unified Catalog?
A) Replace the user's identity. B) Group assets and context for a reusable use case. C) Create a mandatory copy of all data. D) Perform virtual machine backups.
Commented answer
Correct answer: B. The data product brings together assets, purpose, owner, terms, quality, and other information useful for consumption.
Question 3: What does data lineage help to understand?
A) Only those who logged into the portal. B) Origin, movement, transformation, and destination of the data. C) Only the storage price. D) Only retention policies.
Commented answer
Correct answer: B. The lineage supports root cause, impact, audit, and trust in the derived data.
Question 4: Which difference between discovery and access is correct?
A) Finding an asset always grants access to the content. B) Discovery allows you to locate and understand metadata; access remains subject to policies and permissions. C) Access exists only for Data Map administrators. D) Search overrides controls at the source.
Commented answer
Correct answer: B. The catalog can make metadata searchable without automatically releasing the underlying data.
25. Essential Glossary
Term
Meaning
Asset
Individual object described in the catalog, such as a table, file, or report.
Metadata
Information that describes structure, context, operation, or data relationships.
Data Map
Metadata graph that underpins discovery and governance.
Scan
Process of connecting to and capturing metadata from a source.
Ingestion
Processing that populates the Data Map with metadata.
Classification
Category applied to an asset or attribute.
Lineage
Representation of the origin, transformation, and destination of the data.
Unified Catalog
Experience in discovery, curation, and business-oriented governance.
Governance domain
Organizational limit for product and concept ownership and governance.
Data product
Set of assets and context organized for a purpose.
Glossary term
Standardized business term.
Critical Data Element
Logical representation of critical information in different assets.
Data owner
Responsible for decisions and accountability regarding data.
Data steward
Responsible for the daily curation and quality.
Data quality
Adequacy of the data to the purpose.
Data health
View of controls, scores, and governance actions of the assets.
Official references consulted
Microsoft Learn - Study guide for Exam SC-900: Microsoft Security, Compliance, and Identity Fundamentals.
Microsoft Learn - Data governance with Microsoft Purview.
Microsoft Learn - Learn about Microsoft Purview Data Map.
Microsoft Learn - Scans and ingestion in Microsoft Purview Data Map; Scan data sources; scanning best practices.
Microsoft Learn - Data lineage in Microsoft Purview and lineage user guidance.
Microsoft Learn - Learn about Microsoft Purview Unified Catalog.
Microsoft Learn - Governance domains, data products, glossary terms, and search in Unified Catalog.
Microsoft Learn - Data quality, scores, health controls, health actions and reports in Unified Catalog.
Microsoft Learn - Data governance roles and permissions in Microsoft Purview.
Note about update
The Microsoft Purview experience evolves rapidly. The availability of the Unified Catalog, preview features, names, licensing, permissions, and integrations may change. For actual deployment, always check the current official documentation.
