Home Business & Economics Chapter 6 Big Data Governance Framework
Chapter
Licensed
Unlicensed Requires Authentication

Chapter 6 Big Data Governance Framework

Become an author with De Gruyter Brill
Big Data Management
This chapter is in the book Big Data Management
Chapter 6Big Data Governance FrameworkBig governance frameworks are crucial to establishing a successful data governanceprogram across the enterprise. The framework must be scalable, extensible, andyet flexible enough to frame the data governance strategy and implementation.Defining the hierarchy of governance imperatives in the form of a priority pyramidcan clarify and communicate governance priorities.One advantage of viewing big data governance as a pyramid is that it exposesgovernance as a hierarchy and layers of policies. Figure 6.1 shows a four-layer pyra-mid for governance. At the peak of the pyramid, we enjoy the benefits and value ofgovernance by having fully governed and trusted data. This data is available to theentire enterprise for data scientists and data engineers, subject matter experts andprivileged users who can run their data analytics models, business intelligencereports, ad-hoc queries, data discovery, and data insight extraction.The success of the top layers depends on the effectiveness of the lower layers. Thefully governed and trusted enterprise data platform (layer 4) requires a function-ing and effective sandbox for data scientists and the user community (layer 3).Layer 4 is the level where users conduct their modeling, big data analytics, set updata analytics pipelines and data products. Layer 3 is the level where users refinetheir data, process it, cleanse, transform, and curate itprepare it for analysis.EnterprisePlatform(Warehouse)Data scienceworkspaceData lake–Integrated sandboxesRaw, source data–landing area (“full fidelity data”)User community ad-hoc queries, big data analytics, reportsData modeling, data blending, big data analytics pipelines and productsData is refined, processed, cleansed, well defined, completeBroad data collection, raw data1234Data fully governed (trusted)Metadata: Searchable catalogILM: Access control, retention policyQuality: Monitor and test completenessMetadata: Register dataILM: Access control,retention policyFigure 6.1:The Big Data Governance Pyramid.https://doi.org/10.1515/9783110664065-006
© 2020 Walter de Gruyter GmbH, Berlin/Munich/Boston

Chapter 6Big Data Governance FrameworkBig governance frameworks are crucial to establishing a successful data governanceprogram across the enterprise. The framework must be scalable, extensible, andyet flexible enough to frame the data governance strategy and implementation.Defining the hierarchy of governance imperatives in the form of a priority pyramidcan clarify and communicate governance priorities.One advantage of viewing big data governance as a pyramid is that it exposesgovernance as a hierarchy and layers of policies. Figure 6.1 shows a four-layer pyra-mid for governance. At the peak of the pyramid, we enjoy the benefits and value ofgovernance by having fully governed and trusted data. This data is available to theentire enterprise for data scientists and data engineers, subject matter experts andprivileged users who can run their data analytics models, business intelligencereports, ad-hoc queries, data discovery, and data insight extraction.The success of the top layers depends on the effectiveness of the lower layers. Thefully governed and trusted enterprise data platform (layer 4) requires a function-ing and effective sandbox for data scientists and the user community (layer 3).Layer 4 is the level where users conduct their modeling, big data analytics, set updata analytics pipelines and data products. Layer 3 is the level where users refinetheir data, process it, cleanse, transform, and curate itprepare it for analysis.EnterprisePlatform(Warehouse)Data scienceworkspaceData lake–Integrated sandboxesRaw, source data–landing area (“full fidelity data”)User community ad-hoc queries, big data analytics, reportsData modeling, data blending, big data analytics pipelines and productsData is refined, processed, cleansed, well defined, completeBroad data collection, raw data1234Data fully governed (trusted)Metadata: Searchable catalogILM: Access control, retention policyQuality: Monitor and test completenessMetadata: Register dataILM: Access control,retention policyFigure 6.1:The Big Data Governance Pyramid.https://doi.org/10.1515/9783110664065-006
© 2020 Walter de Gruyter GmbH, Berlin/Munich/Boston
Downloaded on 10.10.2025 from https://www.degruyterbrill.com/document/doi/10.1515/9783110664065-006/html?srsltid=AfmBOop-ICxYAdujxSi3zm2a7pISPcGjjy3TS7KPq2HvXm9gleFuYU5i
Scroll to top button