<< 1 >>
Rating: 
Summary: A must read - back to the basics
Review: I've always hated reading techical books that are more than a year old. However, there are a few great ones out there that deserve a read. I have read them all (Kimball, Inmon, etc) this one rates up there with the best of the best.If you want to understand data architecture then buy this one. I like the simple explanations and the logic and common sense approach to this much needed process. Get it and read the first 3 chapters, if you don't like them you won't like the rest of the book. If you don't then return it, but I bet you won't.
Rating: 
Summary: A disappointing read
Review: Michael Brackett's book was a major disappoint to me. My overall recommendation is that it is re-editted or, better still, rewritten. I would advise people to avoid it. My many reasons for this view include: 1) The failure to define the Common Data Architecture (CDA) in explicit terms - what does it actually consist of? The book fails to explain the added value of a CDA over, say, better documentation. 2) The failure to define the Formal Data Resource in practical terms. Is it a data warehouse, a meta data repository or a set of applications? This has a practical impact on its design & build. 3) p87 states that the CDA is not meant to resolve conflict; source system inconsistencies are merely documented in the CDA. This undermines the fundamental benefit of the CDA - consistent and "common" data definitions. 4) Many terms in the book are ill-defined, e.g. what does "organization" refer to on p188, a department, a business unit, an enterprise? Many terms are defined by merely restating the original term, e.g. "The Data Resource Directory component contains the data resource directory" (p161, see also p46, p129 & p140). The author must use alternative words to provide further insight - "a black cat is a cat that is black" doesn't explain what a "cat" is to me. 5) The CDA naming convention allows for attribute names of upwards 20 words long - this is impractical. We are then asked to cross-reference source system names to these nominal names for each screen, file, report etc. - this is unmanageable. 6) The Data Modeling syntax (Chp 5) is very limiting and unnecessary as many well known alternatives exist. Meanwhile, the examples provided show disappointingly bad practice in data modeling. 7) The Physical Data Structure definitions all rely on unsubstantiated assumptions, e.g. "A disparate data item... represents two or more unrelated data characteristics [sic. attributes]" - a data item may combine other attributes, but the vast majority do not. On p226, "Foreign keys are not easily identified in disparate data because they are never defined" - this is simply not true. 8) The author consistently misuses the term "Denormalization" - this is only part of the conversion from Logical to Physical design and not the process itself. 9) Data Quality is often broken down into Accuracy, Completeness, Consistency, Timeliness, Uniqueness & Validity...not just Accuracy, Completeness & Integrity. 10) Finally, chapters 9-13 repeat much of the discussion of chapters 3-8. A better approach would be to combine them to avoid repetition.
Rating: 
Summary: WHAT you will need to do, not so detailed on HOW
Review: This book presents concepts for organizing meta-data and sharing data across an enterprise. It is a strong introduction to general meta-data management and data sharing.
It is not a "cook-book" for implementing a data-wharehouse or data-dictionary project. If you do NOT already have a strong understanding of the basic principles of meta-data organization and data-sharing, I can recommend this book. However, if you are looking for specific "project" plans and steps towards successfully implementing an enterprise data-warehouse or data-dictionary implementation, there are other books that have more specific information on how to do this.
Rating:
Summary: WHAT you will need to do, not so detailed on HOW
Review: This book presents concepts for organizing meta-data and sharing data across an enterprise. It is a strong introduction to general meta-data management and data sharing.
It is not a "cook-book" for implementing a data-wharehouse or data-dictionary project. If you do NOT already have a strong understanding of the basic principles of meta-data organization and data-sharing, I can recommend this book. However, if you are looking for specific "project" plans and steps towards successfully implementing an enterprise data-warehouse or data-dictionary implementation, there are other books that have more specific information on how to do this.
Rating:
Summary: WHAT you will need to do, not so detailed on HOW
Review: This book presents concepts for organizing meta-data and sharing data across an enterprise. It is a strong introduction to general meta-data management and data sharing.
It is not a "cook-book" for implementing a data-wharehouse or data-dictionary project. If you do NOT already have a strong understanding of the basic principles of meta-data organization and data-sharing, I can recommend this book. However, if you are looking for specific "project" plans and steps towards successfully implementing an enterprise data-warehouse or data-dictionary implementation, there are other books that have more specific information on how to do this.
<< 1 >>
| 
