Last updated on 2024年9月21日

How do you create and maintain a data dictionary for your data warehouse?

由人工智能和领英社区提供技术支持

A data dictionary is a document that describes the structure, meaning, and usage of the data elements in your data warehouse. It helps you and your users understand the data sources, definitions, relationships, and quality rules that govern your data. A data glossary is a subset of the data dictionary that provides a common vocabulary and business context for the data terms. Creating and maintaining a data dictionary and a data glossary for your data warehouse is essential for ensuring data quality, consistency, and governance. Here are some steps you can follow to create and maintain your data dictionary and data glossary.

此文章中的业界达人

由社区从 5 条内容中精选。了解更多

Andrew Jones

?? Created data contracts | I build data platforms that reduce risk and drive revenue | Independent Consultant | Author
Kadhirvelu Ratnasabapathi

VP at Seacoast Bank, LinkedIn Top Voice: | Data Governance | Data Quality | Data Management | Data Modeling | Business…
Lindsay Pettai, CKM, A-CSM, CSPO

?? 3x Top LinkedIn Voice | Data ?? Governance Manager | Program Management | Project Management | Change Management…

1 Identify your data sources and scope

The first step is to identify the data sources and the scope of your data warehouse. You need to know where the data comes from, how it is extracted, transformed, and loaded (ETL), and what are the key business processes and domains that the data supports. You can use a data source inventory or a data lineage diagram to document the data sources and their relationships. You also need to define the scope of your data dictionary and data glossary, such as whether they cover the entire data warehouse or only specific subject areas or data models.

添加您的观点

Lindsay Pettai, CKM, A-CSM, CSPO

?? 3x Top LinkedIn Voice | Data ?? Governance Manager | Program Management | Project Management | Change Management PROSCI | Data Governance Leader | Senior Data Governance Manager | Data Governance Program Leader ??
举报内容
Why create a data dictionary for the warehouse? Simple—because without it, navigating your data is like going grocery shopping without a list. You’ll end up with three kinds of chips, no milk, and wonder why your query returns are a mess! Maintaining a data dictionary is like labeling every aisle in the data warehouse—clear definitions, relationships, and sources neatly mapped out. When the CFO asks for “revenue,” you don’t have five versions of it popping up from the marketing, sales, and finance teams like a bad game of Whac-A-Mole. With the dictionary, everyone’s speaking the same language—no more mystery metadata!

已翻译

赞

2 Define your data elements and attributes

The next step is to define the data elements and attributes that make up your data warehouse. A data element is a unit of data that has a name, a type, a format, a length, and a meaning. An attribute is a property or characteristic of a data element, such as a description, a domain, a default value, a constraint, or a metadata tag. You can use a data element catalog or a data model to document the data elements and attributes. You should also assign unique identifiers and naming conventions to the data elements and attributes for consistency and clarity.

添加您的观点

Andrew Jones

?? Created data contracts | I build data platforms that reduce risk and drive revenue | Independent Consultant | Author
举报内容
A data contract can be a great place to define your data elements and attributes. The contract defines what the data is, what to expect from the data (performance, etc) and how to access it. It's the maintainable interface to the data. Having these defined in the data contract, alongside the data, and populated by the data generators, ensure it is always in sync and always reflects the reality. By being more deliberate about how the data is made available, and defining a data contract to go alongside it, you're making it clear to users that this is data they can rely on and build upon with confidence.

已翻译

赞
Kadhirvelu Ratnasabapathi

VP at Seacoast Bank, LinkedIn Top Voice: | Data Governance | Data Quality | Data Management | Data Modeling | Business Analysis | Business Intelligence | Data Warehousing | Data Architecture | Master Data Management
举报内容
While designing the data models for data warehouses, we define the data elements, both in model diagrams and data model design document. The Data Modeling tool, should be able to generate report on the list of entities, attributes, tables, columns, data types, nullability, descriptions, comments, identifiers, default values, formats, lengths, and contexts. This should suffice for the data dictionary creation and maintenance.

已翻译

赞
Merrill Albert

Enterprise Data Leader, Data Governance Officer, Data Thought Leader, Chief Data Officer, Fractional Governance, Data Evangelist, LinkedIn Top Data Governance Voice, creator of #CrimesAgainstData
举报内容
If you use a data modelling tool, your data dictionary is a part of that model. It's easy to keep in sync and it's easy to pull reports on anything you missed documenting. If you choose to keep the data dictionary separate, then you introduce risk that they're out of sync. You can minimize that risk if you regularly check that every data element in the data warehouse is in the data dictionary. It might be something you need to do on a weekly basis, particularly during development. Once in production, you can check less often, but you still need to check if someone put a new data element into production without the required documentation.

已翻译

赞

3 Create your data glossary

The third step is to create your data glossary, which is a list of the key data terms and their business definitions. A data term is a word or phrase that represents a data element or a group of data elements. A business definition is a concise and clear explanation of what the data term means in the business context. You can use a data glossary template or a data catalog tool to create your data glossary. You should also align your data terms and definitions with the business stakeholders and subject matter experts to ensure accuracy and relevance.

添加您的观点

Jonah Egwu, CDMP?

Data Management| Data Governance| Data Quality| Data Privacy| Data Stewardship| Organization Change/Stakeholder Management
举报内容
Creation of data glossary ( business, technical and operational terms) is a good staring point. It requires collaborative efforts from all business and technical units so this can be adopted at the enterprise level and ingested into the meta data solution where the business assets are mapped to the technical assets.

已翻译

赞

4 Link your data glossary to your data dictionary

The fourth step is to link your data glossary to your data dictionary, which is a detailed description of the data elements and attributes in your data warehouse. You can use hyperlinks, references, or metadata tags to link the data terms in your data glossary to the corresponding data elements and attributes in your data dictionary. This way, you can provide a bridge between the business and technical views of your data and enable your users to find and understand the data they need.

添加您的观点

5 Publish and share your data dictionary and data glossary

The fifth step is to publish and share your data dictionary and data glossary with your users and stakeholders. You can use a document management system, a web portal, or a data catalog tool to publish and share your data dictionary and data glossary. You should also ensure that your data dictionary and data glossary are accessible, searchable, and user-friendly. You should also provide feedback mechanisms and communication channels for your users and stakeholders to ask questions, suggest improvements, or report issues.

添加您的观点

6 Update and maintain your data dictionary and data glossary

The final step is to update and maintain your data dictionary and data glossary as your data warehouse evolves. You need to monitor the changes in your data sources, data elements, data terms, and business definitions and reflect them in your data dictionary and data glossary. You also need to review and validate your data dictionary and data glossary regularly to ensure that they are accurate, complete, and consistent. You should also establish roles and responsibilities, policies and procedures, and quality standards for updating and maintaining your data dictionary and data glossary.

添加您的观点

7 Here’s what else to consider

This is a space to share examples, stories, or insights that don’t fit into any of the previous sections. What else would you like to add?

添加您的观点

Metadata

+ 关注

给文章评分

我们借助人工智能创建了此文章。您认为这篇文章怎么样？

很棒不太好

举报此文章

查看全部

How do you create and maintain a data dictionary for your data warehouse?

1

2

3

4

5

6

7

1 Identify your data sources and scope

2 Define your data elements and attributes

3 Create your data glossary

4 Link your data glossary to your data dictionary

5 Publish and share your data dictionary and data glossary

6 Update and maintain your data dictionary and data glossary

7 Here’s what else to consider

Metadata

给文章评分

感谢您的反馈

更多Metadata相关文章

更多相关阅读内容