Metadata Solutions: Using Metamodels, Repositories, XML, and Enterprise Portals to Generate Information on Demand

I remember when data became a specialty in its own right. No longer viewed as simply being supportive of the processes within an organization, data became an asset that led to solid decision making and improved processes. Data, which sometimes had been collected in haphazard ways, in fact, became so valuable that people were trying to logically connect sporadic and isolated data. As the attempts to unify data were taking place, we all realized that its locations, characteristics, definitions, sources, and access were becoming equally important. Hence, the "birth" of metadata.

Associating metadata solely with data does not do it justice. In fact, there are so many aspects to the world of information that metadata needs to embrace each and every one of them. Whether we realize it or not, metadata is already everywhere. All we are missing is an organized view of metadata, despite its origin. Today's world of information needs an associated metadata solution.

Many of us have already learned that metadata is everywhere, but despite this discovery, we have created more of it, in more places. It is time to make an honest assessment of the metadata efforts and begin to focus on metadata as the gateway to all information. To do that we must understand what metadata really is, where it actually comes from, and how to expand its role in the world of automated intelligence.

Few will debate the importance of metadata. It is time, then, for a book that tackles metadata in a way that will clearly lead the reader toward a metadata solution.

Intended Audience

If you have been given the task of managing your organization's information, this book is clearly for you. Likewise, if you are wondering why the data "dictionaries" or "repositories" that have been offered to you just don't seem to serve their intended roles, this book clearly explains what should have been accomplished. Because metadata is, potentially so broad, this book is meant to provide an excellent "backbone" for those who are charged with actually building a full metadata solution. The following are some of the individuals who will benefit the most from this approach.

• Chief information and/or technology officers who are responsible for assessing the metadata situation within their organizations with the objective of beginning a full, practical, metadata solution implementation process
• Business users who have experienced data inaccuracy, lack of available metadata, and a general inability to find the information they need
• Information technology project managers who are responsible for overseeing the design and development of any data-intense application. Examples include a data warehouse, integrated database, decision support application, customer relationship management application, reengineered series of legacy databases, and/or any type of project that requires an assessment of "what is," with the objective of planning a "to be" improvement.
• Data management professionals who are responsible for the administration, standardization, sharing, and organization of corporate data, especially those with previous experiences that resulted in unsuccessful repository or metadata solution implementations
• Developers, especially those who are faced with integrating or analyzing existing corporate applications
• Software vendors that are struggling with a need to supply standard metadata to product consumers while integrating their own software into a metadata accessible result
• Consulting professionals, particularly those who have faced implementation issues at client companies due to the lack of readily available and accurate metadata

This book provides information and content that will enlighten all of these individual groups. The next section describes some aspects of the book that may appear to be quite technical.

How This Book Is Organized

Consider Metadata Solutions: Using Metamodels, Repositories, XML, and Enterprise Portals to Generate Information on Demand as the first book to address the metadata situation from the beginning through to a practical solution and then into its future maintenance and enhancement. As such, the book has been divided into six parts.

Part I, Today's Information, prepares readers for the book's subsequent discussion of metadata. By looking at information in a way that many may never have considered, this part provides an overview of information and its many existing perspectives. Then, the information problems that have surfaced are discussed. Finally, this part reviews information solutions that have been tried, and for the most part are still in existence, and explains where they fell short. At the conclusion of Part I, the reader should be ready to discuss metadata. Many experienced information practitioners can skip this part, but even with information integration experience, some eyes may still be opened.

Part II, Metadata as Part of the Solution, begins the discussion of metadata itself. First, a solid definition of today's buzzword focuses readers on how tunnel vision can even affect cross-application concepts such as metadata. The part begins the metadata requirements process. By stepping you through a methodology that first identifies metadata beneficiaries, determines metadata requirements, and then begins a categorization process, Part II gets the reader used to metamodels. Metadata stores--the physical storage locations for metadata--are also discussed, giving you various options as to how metamodels can be implemented. As a way of reminding the reader that metadata solutions are much more than the storage of metamodels, the metadata solution architecture is the last topic discussed in the part.

Part III, Entering Meta-Meta Land, takes us inside the metadata solution. To deal with the fact that metadata is everywhere, a true metadata solution needs to be cognizant of the location and access requirements of existing metadata. In addition, metadata solutions all process and display their metadata differently based on the type of metadata. Designers and developers of true metadata solutions must be able to treat metadata with a software perspective. Part III focuses on what metadata means to a tool and discusses the meta-metamodel. Once this basic understanding is covered, metadata-based technologies, such as repositories, the Web, XML, and file management systems, are all discussed. Some aspects of the part may be too technical for the casual reader, but the chapters do clearly explain the internals of metadata solution technology.

Part IV, Beginning the Metadata Solution Process, discusses implementation-specific aspects, other than the metadata and its associated metamodels. Nontechnical factors, such as readiness, scoping, and internal environment changes, are addressed at the beginning of this part. The discussion then moves to technical factors--multitool architecture, metadata update and exchange, metadata presentation. A chapter is dedicated to metadata solution technical support, specifically metadata and repository administration. Part IV ends with advice on determining the right solution.

Part V, Sample Metadata Solutions, begins with a case study--A Typical Metadata Disaster--that equates to a very common metadata situation within corporate America. Succeeding chapters illustrate actual metadata solution implementations that are all focused on solving the identified disaster. Illustrated solutions (often including actual program code, metamodels, and architectural diagrams) include a centralized repository, an integrated repository architecture, an information directory, metadata-interexchange using XML, a standalone metadata store, and an enterprise portal. Although some aspects of this part are quite technical, I strongly urge all readers to at least browse the various solutions.

Part VI, Maintaining the Metadata Solution, describes how the metadata solution stays alive. One way is by ensuring that it meets the requirements of its targeted beneficiaries. This final part discusses the organizational responsibilities that go along with such a task; it also focuses on how to ensure the livelihood of the metadata itself by discussing metadata quality. The book closes by pointing out where metadata meets the business strategy, now and into the future.

Reading Paths

Implementing a metadata solution involves a variety of skills, from business analysis all the way through to technical application and interface development. As such, the chapters in this book range in terms of primary audience and interest level. Those chapters that are geared purely to those responsible for hands-on metadata solution delivery because they contain sample code or metadata solution internals have been labeled Technical at the upper right corner of the first page in each chapter.

To accommodate the different backgrounds of this book's readers, I have set up the following reader categories and noted the chapters that will be of interest.

• Information systems management people who need to be aware of the intricacies of metadata solutions, but have not planned to have any hands-on involvement:

  − Part 1 - all chapters

  − Part 2 - all chapters

  − Part 3 - Chapters 15 and 16

  − Part 4 - all chapters

  − Part 5 - Chapters 21, 24, and 25

  − Part 6 - all chapters

  • Business users who crave a well-implemented metadata solution

    − Part 1 - all chapters

    − Part 2 - Chapters 7, 8, and 9

    − Part 3 - Chapters 15 and 16

    − Part 4 - Chapters 17 and 20

    − Part 5 - Chapter 21

    − Part 6 - all chapters

    • Technical analysts, and developers who are familiar with database technology

      − Part 1 - Chapters 1, 2, and 6

      − Part 2 - all chapters

      − Part 3 - all chapters

      − Part 4 - all chapters

      − Part 5 - all chapters

      − Part 6 - all chapters

      • Data management professionals who are familiar with metadata and its current treatment

        − Part 1 - Chapters 1, 4, 5, and 6

        − Part 2 - all chapters

        − Part 3 - Chapter 12, 14, 15, and 16

        − Part 4 - all chapters

        − Part 5 - all chapters

        − Part 6 - all chapters

      Model Legend

      Throughout the book, various models are used to illustrate metadata relationships, metamodels, and metadata flows. Because, unfortunately, there really is no uniform way of depicting models these days, a figure that shows the various symbols you will see in this book will be printed on the inside cover.

      Most of the illustrations in this book follow my default modeling characteristics, at the top of the figure. However, because other illustrations and models have been brought in from other sources, it is important to understand their notations, as depicted. The Unified Modeling Language (UML) is the standard followed by the Object Management Group (OMG).

      What Is Your Objective?

      As you read this book, you should be correlating the described metadata situations with those of your own. Actual case studies, submitted by large organizations that have lived through metadata-related situations, are used throughout the book. You should consider how a renewed metadata perspective, like the one discussed throughout this book, can revitalize the metadata that exists at your organization. Readers should expect to gain enough knowledge to move forward on a hands-on metadata solution implementation of any scope.

      0201719762P04172001