Conference Session Abstracts

Monday February 27, 2006

Data Management Maturity - Assessing Your Environment
Peter Aiken
Data Blue Print

Organizations recognize the SEI/CMMI process-improvement framework's ability help to improve complex processes. Many recognize that mature data management (DM) practices are a necessary but insufficient prerequisite to using data effectively as an organizational asset. That is, DM must be practiced organizationally with the rigor required by other advanced professions (i.e., software, building, aircraft, accounting, and legal). In this presentation, Peter Aiken will discuss the results of an assessment of over 150 organizational DM practices provides insight into various stages of DM maturity; areas of weak, average, and strong DM practice; and implications for you and your organization and the DM community.

Leveraging Common Metadata for Data Synthesis AND Transformation!
David Friedland, Rob Howard
IRI

Sometimes operational data are not available to the application developer, or are kept secure for confidentiality reasons. In such cases, a record-generation and file synthesis solution can be used to emulate the look and feel of real data for the purposes of stress-testing existing applications, designing new applications, and benchmarking. A robust and portable metadata syntax helps define the content and layout of data to be generated, so the created fields, records, and files can be as realistic as possible in terms of size, position, data types, formatting, etc.

But more importantly, when real data become available, the same metadata that were defined at the synthesis stage can then be used for transformations by the data processing tool that leverages that metadata. Conversely, metadata used in transformation applications can be used by the synthesis tool when operational data is either unavailable or confidential -- allowing the synthesis of real-looking (but not real) data and files to begin immediately.

This presentation will focus on the benefits of this shared metadata approach. It will describe through examples how IRI's data synthesis product, RowGen, and data transformation product, CoSORT, use the same metadata and repositories.

Business Metadata Strategies
Bonnie O'Neil
Project Performance Corp

This tutorial/workshop describes the importance of business metadata and the benefits it provides to the business if it is managed properly. Business metadata has as its main focus presenting contextual information to business users in their own language, and helps them interpret data so they can understand it better. Business metadata is often neglected even in sophisticated metadata efforts, and this session explains why and what you can do to avoid neglecting it. The tutorial explains the main subject areas that business metadata includes. Practical techniques are presented that enable attendees to create a plan to implement business metadata in their environment.
Other areas covered include:
- The role of lexicons
- The role of tools
- Taxonomies and Ontologies
- Business Rules
- Data Quality and Business Metadata

Tuesday February 28, 2006

Dictionary on a Shoestring!
Bonnie O'Neil
Project Performance Corp

This presentation is about launching a dictionary initiative on a shoestring, using Bonnie's Law: Use Whatever Technology is Lying Around. The chronicles of a media company that went through a major migration will be presented, illustrating how they were left with all sorts of conflicting business terms (not to mention data elements too!), and how they solved the problem, with no budget at all. Lessons learned from the exercise include:

* How to launch a business metadata initiative when your client doesn't know what that is
* What the organization gains by this initiative
* Basic shoestring principles: How to launch all sorts of metadata initiatives with no budget
* Why dictionaries are important

Controversial Issues in Data Modeling
Tom Haughey
InfoModel

In this session, modeling guru, Tom Haughey will focus on the most controversial issues discussed and debated by practicing data modelers.

Highlights include:
* Conceptual vs Logical Data Models
* Data Modeling - analysis or design?
* Handling Domains and codes tables

Creating Metadata from Data: A New Approach to Metadata Management
Denise Sparks
Exeros

While there has been considerable discussion about the advantages of metadata management solutions after they have been deployed, there is little talk about how to discover, document and enter metadata into your repository. The reason this topic is avoided is that documenting metadata for new data sources and the relationships between data sources is a manual, laborious, risky and expensive proposition, often involving thousands of data analyst, business analyst and engineering hours. So how do you get from where you are today to the point of reaping those benefits? Attend this session and learn how new technology approaches can now examine data and derive metadata not only about a single data source, but can automatically discover complex metadata maps and transformations between data sources and can help you easily populate metadata repositories keep them up to date.

Attend this presentation to learn how the Exeros's innovative data-driven technology has resulted in 5X reduction in discovering and documenting metadata within and between data sources at major financial services institutions.

Data Modeling Case Study

Abstract will be posted shortly

Taxonomy Derivation: The Nuts and Bolts of Actual Derivation
Seth Earley
Earley and Associates

What are the steps in deriving a taxonomy? Where do you start? What
questions do you ask? What are the best sources for terms? How are they
arranged into the actual taxonomy? In this session, you will learn the
tactics of taxonomy derivation-the specific process steps that you need to go through to get to your end result. The session leader will discuss user interviews, types of questions to ask, working session techniques, work task analysis, content analysis, how taxonomy terms can be pulled from process maps, ways of understanding patterns and themes, and how to think about metadata fields versus taxonomy term values. You will leave this session with a clear understanding of how to derive your taxonomy

Data Model Patterns: A Metadata Map
Dave Hay
Essential Strategies

While there have been numerous books written about the advantages of addressing metadata and even about how to go about it, precious little has been written on what exactly the stuff is. This is a pity, since most of the controversies that have plagued our industry for the last 10 years can easily be put to rest with proper understanding of the metadata perspective on them.

David Hay is the creator of a comprehensive "enterprise data model" of the systems development world, and as such provides a scheme for an idealized metadata repository. It is organized in terms of John Zachman's Architecture for Enterprise Development, which means that it goes beyond just data.

The complete model also describes process modelling, people and organizations, locations, timing and events, and motivation. It addresses the perspectives of business users, architects, and system designers.

This presentation is of the portion of that model that concerns the Data column, across the Business Owner's, Architects, and Designer's views, along with that of the Functioning System. The presentation is based on the author's forthcoming book from Morgan Kaufmann Publishers, Data Model Patterns: A Metadata Map.
The paper will describe techniques for dealing with variable attribute entities. It will discuss the implications of this approach on mapping to legacy systems, and how it actually is a model of relational theory itself.

Metadata Repositories: Build or Buy?
Malcolm Chisholm
AskGet Inc.

Enterprises face increasing needs to manage their information assets, and are coming to realize that part of the solution is metadata management. Repositories are databases that manage metadata, and so the obvious answer is to implement a repository, either by buying or building one. Metadata is often thought of as some pre-defined set of information that varies little from one organization to another, and so buying a repository can answer the fundamental information management issues that the enterprise is trying to address. There is some truth in this, but it is not the whole picture. Major problems can arise when an enterprise has specific metadata management requirements that do not fit well with generic repository functionality. Indeed, there can be wide variations in such requirements from enterprise to enterprise. Building a repository versus buying one is therefore an important question that must be answered early in the process of developing a metadata strategy.

This session will discuss:
- The basic structures of repositories and common classes of metadata that they manage.
- The wide variations in requirements for metadata management that exist across enterprises
- Defining repository requirements using use cases
- Managing the process to buy a repository
- Managing the process to build a repository.

Questing for the Grail: The post-mortem of a federated metadata management project.
William Brooks
Data and Integration Architect
MFS Investment Management

Beginning in late 2002, MFS began developing a strategy for integrating and managing a wide variety of business and technical metadata. The centralized project team, within the company's IT group, sought to develop an approach to store and provide access to metadata from physical databases, an XML-based messaging infrastructure, ETL tools, data modeling tools, and enterprise scheduling systems. The project accomplished many of its goals, but still fell far short of the "holy grail" it had intended to produce: a unified, universal Metadata Repository. Although not a stunning success, the project was far from a failure. Bill will explain the nature of the project, why it didn't meet its original lofty goals, and the lessons that MFS has been able to integrate into its subsequent metadata (and data) management approach.

Attendees will learn:
· About the design and architecture decision process for a company-wide federated metadata management strategy.
· The nature of the data environment and how the approach was chosen.
· Methodologies used to organize federated metadata in the project.
· The outcomes of the project and lessons learned

Panel -What Practitioners Want from Metadata Tools
What Practitioners Want from Metadata Tools

Are you getting the optimal use from your metadata tool? Is the metadata tool market confusing to you? Are there changes you would like to see implemented by tool vendors? Additional functionality? Easier to use?

In this interactive panel session moderated by Peter Aiken, practitioners will discuss their experiences and recommendations regarding metadata tools. Bring your questions and suggestions to this interactive session.

Topics discussed will include

- Tool evaluations and assessments
- Is there really an enterprise metadata tool?
- Effectiveness of homegrown metadata tools
- Building the business case for tool purchase
- Getting the most out of your metadata tool
- Tips to sell and market use of your metadata tool
- Ease of use of tools on the current market
- Standards for metadata tools
- Interoperability between tools
- Metadata integration
- Recommendations and suggestions for future requirements

Wednesday, March 1, 2006

Semantic and Contextual Extraction
The Business Value of Untangling Loosely Structured Data
David Rafner
Data Blueprint

A growing number of operational systems are awash in loosely structured data that is hard to search, analyze, or migrate. Have you been confronted with the challenge of tapping these systems but found your traditional tools and tricks lacking? Indeed, it is rare to meet a data professional who has not had to struggle with reports, contact logs, emails, or weakly-typed customer and product data of dubious quality. Improving internal and industry standards are part of the solution. But the variety and complexity of the information ensures that this problem will not go away anytime soon.
Fortunately, there are now tools and methods that can help you begin to untangle and enhance the business value of this data. Beyond clever pattern matching, semantic and contextual analysis can harness your subject matter expertise to identify business data structures. The discussion will cover techniques for restructuring into XML and how heuristics, ontologies, and taxonomies impact the extraction process. The positive business impact is demonstrated with an eye toward both user access and long-term data quality.


Master Data Management
Versus
Reference Data Management
Malcolm Chisholm
AskGet Inc

There is an increasing emphasis on the need to manage master data in enterprises. However, the relationship between master data and other classes of data are not always clearly understood. This is especially true of the way that master data interacts with reference data. Both master and reference data have unique management needs. If these needs are not addressed, it is unlikely that an enterprise will be able to effectively manage either class of data.

This presentation discusses the characteristics that define master data and reference data. The management needs of these classes of data are described, with emphasis on the differences between master and reference data. However, master data often has a life cycle that inevitably involves reference data, so there needs to be coordination between the strategies for managing the two classes of data. Trade-offs between local versus central management are also discussed.

The following topics will be covered:

- Definitions of master data, reference data and other classes of data
- Characteristics of master data and reference data
- Management strategies for master data contrasted with those for reference data
- Integration between master and reference data
- Common problems in managing master and reference data

Metadata and the Law
Did You Think Metadata Only Matters to IT?
Conrad Jacoby, Esq.
General Counsel
Potomac Consulting Group

Computer metadata has become a prized resource for lawyers engaged in civil litigation. Metadata is used to help identify key players in legal disputes, uncover deleted language that shows how decisions may have been reached, and is important to demonstrating that electronic materials are authentic and have not been revised or forged by unscrupulous parties. This presentation showcases different ways that metadata has become important to the legal community and provides examples of corporate metadata important to in-house legal departments because of litigation and regulatory compliance systems such as the Sarbanes-Oxley Act.

Advanced Data Model Patterns
Dave Hay
Essential Strategies

The book Data Model Patterns: Conventions of Thought describes a set of standard data models that can be applied to standard business situations. These patterns, it turns out, occur on several levels. At the basic level are models of the things seen in business. The patterns in the book are a bit more abstract than conventionally seen, but they do describe things that are easily recognizable to anyone: people and organizations, products, contracts, and so forth.
There is a more abstract level of modeling, however, which is necessary when the things being modeled don't fall into these tidy categories. This level, also described in the book, is the subject of this presentation.

Enterprise Architecture Trends and Regulatory Audit Ability
Service Oriented Architecture
Don Soulsby
MetaWright

The Service Oriented Architecture (SOA) is presented as a solution to an enterprise's need to manage the seemingly chaotic collection of applications, interfaces and information formats that are presented to the business consumer. The SOA approach delivers both operational and business intelligence applications through a set of standard components delivered to the client via web services.
SOA is primarily focused on the delivery aspects of an enterprise information architecture. The lines between multiple applications and data sources will be blurred with the delivery of services that are tuned to specific business processes. With the increased focus of an enterprise on governance, liability and risk management, the need for clarity and traceability from the SOA services to the component and primitive models will become critical. On of the important aspects of the architecture is the collection and integration of metadata. The presentation will look at a practical approach to managing the SOA metadata within the context of a complete enterprise architecture with particular emphasis on the delivery of business intelligence components.

Mapping Your Enterprise Data Architecture Strategy
Greg Keller
Embarcadero Technologies, Inc.

According to CIO Magazine's "The State of the CIO 2004" survey, integrating and enhancing both systems and processes - as well as ensuring data integrity - are the two leading priorities of CIOs and IT executives in 2005.
And when it's their priority, it's your priority.
To make matters worse, if these challenges aren't addressed properly, you risk becoming the focal point of poor business decisions based on inaccurate or irrelevant data.
Whether it's a business intelligence implementation or a mandate to streamline business processes, a well-thought-out approach to data management becomes key to your success. But when you delve into your applications, it's easy to become lost in the wilds of your corporate data.
That's why organizations are taking a model-driven, metadata-rich approach to building an enterprise data architecture strategy. Using metadata effectively can help you clearly document the context of your data, ensuring that the data you deliver is accurate and relevant.
Learn how you can leverage modeling and metadata to:
* Ensure data quality - understand how metadata can help you bridge the gap between how data is stored in disparate applications
* Intelligently integrate information across the enterprise - see how rich information and standardization can save you time integrating your applications
* Leverage existing data assets - make sure that you use your data assets effectively by providing more information about information you already have

Metadata Requirements for Regulatory Compliance and Information Security
David Schlesinger
Intel

This presentation will describe the two major levels of information characterization that must take place to understand the constraints demanded by numerous data regulations, as well as to protect your company's intellectual assets. Failure to break your information characterization into separate components for categorization makes subsequent evaluation and regulatory adjustment much more difficult and costly.

The talk will show the two basic information definition levels, explain the major data regulatory compliance categories, and include a discussion of when in the process of creating information systems and handing out information to employees this activity needs to occur.

Boosting the Bottom Line with Master Data Management
John McAllister
VP Client Services
Stratature

Companies around the world are immediately realizing the benefits of master data management (MDM). MDM centralizes key data and data structures-concerning customers, products, accounts, and other business dimensions-across an enterprise's business intelligence, data warehouse, ERP, financial and operational systems.

MDM is helping companies dramatically increase productivity, improve data quality, ensure data integrity, enhance corporate compliance, and maximize their return on investment from both current and future technologies.

This session takes an in-depth look at real-world case studies in master data management. RSM McGladrey Employer Services (a subsidiary of H&R Block) ties customer contract master data into their billing system to immediately increase customer billings by $375,000/month - that's $4.5 million in the first year alone. Tiger Brands, an international packaged goods company integrates customer data from 23 separate customer systems to reduce frequent hierarchy updates from 161 hours to one hour. Compass Bancshares applies MDM for cost center management and chart of account standardization to reduce dimension maintenance of seven source systems (business intelligence/BPM, general ledger, costing, risk, etc.) by over two weeks per month.

SPECIAL SESSION

What's inside the "Black Box"? Leveraging subject area and logical data models to understand ERP systems
Steve Hoberman
Steve Hoberman and Associates

Enterprise Resource Planning (ERP) applications such as SAP, Siebel, and PeopleSoft have been implemented and run in many organizations worldwide. These applications are promoted as self-contained systems (informally known as "Black Boxes" which for the most part require only a basic knowledge of how to enter data into the system and transfer data in and out of the system using standard interfaces. The concept of self-contained is a strong selling point for these applications, as corporations can save time and money by avoiding the "reinvent the wheel" concept. However, situations arise where we need a detailed level of understanding of the structures within these systems. This presentation will discuss how I used a series of data models to explain a very complex area within SAP addressing my approach and the benefits derived.

Lessons Learned Metadata, Data Modeling, Compliance, Data Security
So Where Are We Really Going?
Peter Aiken, Don Soulsby, David Schlesinger, Intel, Mike Ley, UPS, Steve
Hoberman, Mars

 

© Copyright 2004 DebTech International, LLC. All Rights Reserved.