Rating: 
Summary: Very good, but has some worrying errors
Review: This is a very good book about metadata, but I can only give it 3 stars because the author makes some very worrying (and sometimes quite amusing) mistakes.
For example, in the preface page xxii, the UML semantics for unrefined association, aggregation and generalization are *completely* wrong. So is the syntax for UML object. This is a very worrying error for a metadata book. Finally, there is the amusing quote on page 180, "if you have mastered the art of metadata by mastering the ability to think like a tool...". In England, "tool" is actually a well-known synonym for d*rk! Not the author's fault, but one would expect an editor to pick up on something like this!
Apart from these unfortunate errors, the book is well-written and very readable, and will certainly give you lots of useful information and techniques for working with metadata. Definitely one to keep on your bookshelf for reference. Well done Adrienne!
Rating: 
Summary: The Most Hands-On around
Review: Those of us that have to implement something will really appreciate this book. THere are others on the market that introduce us to metadata concepts, but leave the implementation to vendors. Some of us know that this can fall short.If you already have a vendor product and want to improve your situation, consider this book as a starting point. It covers a realistic definition of metadata in terms of not just data, but the world around data (programs, files, business rules, physical database catalogs, for example). Metadata Solutions also shows the reader how all of these perspectives are necessary in order to be successful. Metadata can take on two flavors in today's world. The popular view uses it as a detailed description of data warehouse data. Ms Tannenbaum does not necessarily fault that viewpoint, but shows us how to some degree we are participating in a potential disaster by keeping copies of metadata that may already exist in a standalone data warehouse product. Likewise, a centralized metadata repository might also create the same problem. Ms. Tannenbaum likes the portal approach, but she does show examples of virtually all metadata solutions, from separate standalone database, through XML transfer, through a fully integrated portal-based "search and retrieve" solution. Is this for managers? Yes...but they don't need to read the whole thing..(in fact she even gives reading paths). Is this for data people? Definitely. Is this for developers? definitely. But most important, it is for anyone who thinks they know what metadata is about....surprise!
Rating:
Summary: More and more relevant
Review: We have been struggling with metadata since our first data warehouse was built. Lots of problems have been slowing this process down. First, the definitions - not as easy as you think. We are still struggling with a "Customer" definition. Next, the source..some of our data results from many sources, some doesn't have a source. So how do we depict this? Everyone has an answer. I started looking to books. Read data quality books, data warehousing books. Only this book takes the "metadata problem" approach. The book is broken down into several parts. I thought some of it was beyond my understanding at first. I discussed some of the book's ideas with technical developers here, and when I went back to Part 4 (meta-meta land) and Part 5 (Sample Metadata Solutions), I understood what needs to be done in order to solve our metadata problems. I am not sure if we will ever come to terms on what is right and what is wrong, but with a "metadata solution", everyone can be happy before this decision is made. I recommend this book to everyone that has been struggling with existing metadata. Consider the methodology depicted in this book as a way to sift through all of it. Good book, and I keep going back to it as we plan our new metadata solution approach.
| 
