Metadata strategy must be aligned with business priorities

Ensure future scalability from an enterprise perspective, achieved through iterative and incremental delivery

Impressing its value will encourage businesses to use metadata while providing knowledge assistance to businesses.

Ensure employees understand how to access and use metadata

Metadata is usually generated through existing processes (data modeling, SDLC, business process definition), so the process owner is responsible for the quality of the metadata

Develop, implement and review metadata standards to simplify the integration and use of metadata

Create a feedback mechanism so that data users can report incorrect or outdated metadata to the metadata management team

Focuses primarily on the content and conditions of the data, but also includes details related to data governance

Provides information about the technical details of the data, the systems in which the data is stored, and the processes by which data flows within and between systems

Describes the details of processing and accessing data

Description metadata

Structural metadata

Manage metadata

bibliographic metadata

Record metadata

Save metadata

Metadata repository refers to the physical tables that store metadata, often built into modeling tools, BI tools, and other applications

The purpose of a business glossary is to record and store an organization's business concepts, terms, definitions, and the relationships between these terms

As with all data-oriented systems, designing a business glossary should consider hardware, software, databases, processes, and human resources with different roles and responsibilities

The Business Glossary application needs to be built to meet the functional needs of three core users

business user

technical user

Data Management Specialist

Business intelligence tools generate various metadata related to business intelligence design

A configuration management tool or database CMDB provides the functionality to manage and maintain metadata related to IT assets, their relationships, and the contract details of the assets.

A data dictionary defines the structure and content of a data set, typically for a single database, application, or data warehouse

There is one data dictionary for each database, and the data dictionary in each database is not universal.

Many data integration tools use executables to move data from one system to another, or between different modules within the same system

Database catalogs are an important source of metadata. They describe the contents of the database, information size, software version, and other operational metadata attributes.

The most common form of database is relational, which manages data as a set of tables and columns

Metadata solutions should be able to connect to various databases and datasets and read all metadata exposed by the database

Mapping management tools are used during the analysis and design phases of projects, converting requirements into mapping specifications, which are then used directly by data integration tools or used by developers to generate data integration code

Data quality tools assess data quality by validating plans

Data dictionaries and glossaries contain detailed information about terms, tables, and fields, but dictionaries or catalogs contain information about the systems, sources, and locations of relevant organizational data.

Event messaging tools move data between different systems, require extensive metadata, and generate metadata that describes the movement

Data modeling tools are used to build various types of data models: conceptual, logical, and physical models

Reference data records the business value and description of various types of enumerated data for use within the context of the system

A centralized metadata architecture consists of a single metadata repository that includes copies of metadata from separate sources

Organizations with limited IT resources, or those seeking to automate as much as possible, may choose to avoid this architectural option

Organizations seeking a high degree of consistency in a public metadata repository can benefit from a centralized metadata architecture

advantage

shortcoming

process

A fully distributed architecture maintains a single access point

Metadata search engines respond to user requests by retrieving data from source systems in real time

Distributed metadata architecture without persistent database

advantage

shortcoming

process

Hybrid architecture combines the characteristics of centralized and distributed architectures, metadata still moves directly from the source system to the centralized repository, but the repository design only considers user-added metadata, important standardized metadata, and added through manual sources Metadata

The architecture benefits from near real-time retrieval and augmentation of metadata from the source to best serve user needs when needed.

Based on user priorities and requirements, metadata is used as up-to-date and efficient as possible

Hybrid architecture does not improve system availability

Beneficial for organizations with rapidly changing operational metadata that require a consistent, unified metadata organization, and where metadata and metadata sources are growing substantially

For organizations with mostly static metadata or small metadata increments, their potential may not be realized.

It allows metadata to be changed in any part of the schema (source, data integration, user interface) and then synchronizes the changes from the repository (agent) to its original source to enable feedback

dedicated interface

semi-private interface