"A Data Fabric1 has its focus more on the architectural underpinning, technical capabilities, and intelligent analysis to produce active metadata supporting a smarter, AI-infused system2 to orchestrate various data integration styles, enabling trusted and reusable data in a hybrid cloud landscape to be consumed by humans, applications, or other downstream systems. Data cataloging to generate and leverage active metadata is seen as a vital component of any Data Fabric." (Eberhard Hechler et al, "Data Fabric and Data Mesh Approaches with AI", 2023)
"A Data Mesh3 views data primarily as organized around domain owners who create business-focused data products, which can be aggregated and consumed across distributed consumers, organizations, and Line of Business (LoBs) in a self-service and shopping-for-data fashion. Transforming data from disparate data sources to be consumed as data-as-a-product is an essential paradigm of any Data Mesh." (Eberhard Hechler et al, "Data Fabric and Data Mesh Approaches with AI", 2023)
"A data product is based on semantically related raw data that is transformed into a meaningful business context and easily discoverable and consumable by business users." (Eberhard Hechler et al, "Data Fabric and Data Mesh Approaches with AI", 2023)
"An enterprise data warehouse is a central repository of integrated and transformed, structured data from disparate sources and used for reporting and data analysis." (Eberhard Hechler et al, "Data Fabric and Data Mesh Approaches with AI", 2023)
"Data Fabric is an integrated layer of data sources and connection processes based on active metadata." (Eberhard Hechler et al, "Data Fabric and Data Mesh Approaches with AI", 2023)
"Definition of data and AI governance policies, rules, and classifications is critical to break down data silos, allow for a uniform data consumption, and prevent misuse of data. It includes monitoring of compliance and enforcement of data and AI rules and policies on an ongoing basis, as well as ensuring compliance with regulations and laws." (Eberhard Hechler et al, "Data Fabric and Data Mesh Approaches with AI", 2023)
"Drift measures the drop in accuracy and drop in data consistency by comparing accuracy during runtime with the accuracy during training and by comparing key characteristics of the dataset used for training with the dataset during runtime."(Eberhard Hechler et al, "Data Fabric and Data Mesh Approaches with AI", 2023)
"In essence, a Data Mesh solution organizes data around business domain owners and transforms relevant data assets (data sources) to data products that can be consumed by distributed business users from various business domains or functions. These data products are created, governed, and used in an autonomous, decentralized, and self-service manner. Self-service capabilities, which we have already referenced as a Data Fabric capability, enable business organizations to entertain a data marketplace with shopping-for-data characteristics." (Eberhard Hechler et al, "Data Fabric and Data Mesh Approaches with AI", 2023)
"[...] it is the Data Fabric architecture that enables the Data Mesh. In other words, the Data Fabric is the architectural underpinning to implement a Data Mesh solution." (Eberhard Hechler et al, "Data Fabric and Data Mesh Approaches with AI", 2023)
"Semantic enrichment19 is the process of adding meaning to data, which is represented as additional metadata in the knowledge catalog. The intent of semantic enrichment is to simplify and optimize some of the key Data Fabric and Data Mesh tasks, such as search and discovery of assets, access, and consumption of assets by applications and business users to build corresponding data products." (Eberhard Hechler et al, "Data Fabric and Data Mesh Approaches with AI", 2023)
"The aim of a Data Mesh solution is to establish a data marketplace where data can be searched for, discovered, and consumed as a product."
"The goal of semantic enrichment is to simplify and optimize some of the key Data Fabric and Data Mesh tasks, such as search and discovery of assets, access, and consumption of assets by applications and business users."
"The terms Data Fabric and Data Mesh are often viewed as different, conflicting, or at the best overlapping data architectures or frameworks, data management concepts, or approaches to discover, explore, govern, and consume data. However, these concepts are related to each other, where each concept emphasizes specific imperatives or objectives."
"The term data governance3 is used for the processes and responsibilities that define, manage, and enforce access, privacy, availability, and security of the organization’s data. It typically includes a set of policies, rules, and data classifications and functionality to monitor and enforce compliance. As stated earlier, we use the term AI governance in a broader sense, also including AI artefacts."
"The value of a Data Mesh solution is that it assigns the creation of data products to data engineers and subject matter experts upstream who are most familiar with the business domains and corresponding needs."
No comments:
Post a Comment