Track Sessions

DAMA INTERNATIONAL SYMPOSIUM and WILSHIRE META-DATA CONFERENCE
April 28-May 2, 2002 – San Antonio Convention Center, San Antonio, Texas
Agenda is subject to change.

Communication between data modelers and object modelers is often difficult. One of the reasons for this is different terms (“lingo”). In reality, many of the class diagram-related terms have corresponding terms in the data modeling world. This presentation will present UML modeling terms from a data modeler's perspective. Once we have a foundation of terms, we will compare data models to the corresponding class diagram models. For example, given a many-to-many relationship: How is this shown in the class diagram? In the relational data model? We will also study inheritance notation in UML and it’s impact on super/sub-types in the relational data model.

Carrie Sherier has worked in the data modeling field for over 13 years. The first ten years were for a natural gas pipeline company, spending the last few years designing the data model for a major facility management application. She currently works for a major telecommunications company as a Data Architect. Carrie has been working on object-relational projects, utilizing class diagrams to design objects, and implementing them in a relational database. She has been involved in leading the company's data modelers as they make the paradigm switch to object modeling.

From Entities to Stars, Snowflakes, Clusters, Constellations and Galaxies: A Methodology for Data Warehouse Design

Daniel Moody
Associate Professor, Norwegian University of Science and Technology and
Senior Research Fellow, Monash University

This presentation describes a method for developing dimensional models from traditional Entity Relationship models. This can be used to design data warehouses and data marts based on an enterprise data models. The advantage of this is that it provides a more structured design procedure, which is based on the underlying relationships among the data. This allows data marts to be developed in an architected manner and simplifies extraction from production systems. The first step of the method involves identifying “event” entities in the data model¾these correspond to fact tables in dimensional models. The second step involves identifying hierarchies in the model¾these correspond to dimensions in dimensional models. The final step involves collapsing these hierarchies and aggregating transaction data to form dimensional models. A number of design alternatives are presented, including a flat schema, a terraced schema, a star schema and a snowflake schema. We also define a new type of schema called a star cluster schema. This is a restricted form of snowflake schema, which minimizes the number of tables while avoiding overlap between different dimensional hierarchies. Individual schemas can be collected together to form constellations or galaxies. The method is illustrated using a simple example, and also with a real world case study.

Daniel Moody holds a joint position as Associate Professor at the Norwegian University of Science and Technology and Senior Research Fellow in the School of Business Systems at Monash University. He is the National President of the Data Management Association (DAMA) and Australian World-Wide Representative for the Information Resource Management Association (IRMA). Daniel has held senior data management positions in some of Australia’s largest commercial organizations, and has consulted in a wide range of organizations both in Australia and overseas, including Canada, Singapore, Hong Kong, Indonesia, Taiwan and South Korea. He has held academic positions at a number of Australia’s leading universities, including the University of Melbourne, the University of New South Wales, the University of Queensland and Queensland University of Technology. His research interests include data modeling, information resource management, information economics, data warehousing and knowledge management. He has published over 50 papers in the IS field, in both practitioner and academic forums, and has chaired a number of national and international conferences.

Throughout my career, I’ve been plagued with a naïve, idealistic, assumptive approach to implementing programs that will enhance our ability to manage data effectively. After all, who wouldn’t want to support this lofty goal? Each effort rewards me with valuable lessons learned. My recent attempt to define and implement a distributed data stewardship program within a large organization reinforced prior lessons learned, provided some unexpected “gotcha’s”, and established in my mind some specific prerequisites before I would embark on this endeavor again. This presentation is intended to save other zealous data management practitioners from experiencing my pain and perhaps offer an opportunity to achieve more gain.

Carol Knight is an independent data management consultant. She has over 20 years experience in a variety of roles within the data management arena including implementing data management functions; managing DBA and DA staff; evaluating data management tools; creating logical data models; facilitating JAD sessions; designing data management programs, policies, and procedures; and applying methodology to processes. Carol has been a conference speaker on data management topics and an avid supporter of the Atlanta chapter of DAMA.

Business Intelligence (BI) is not a project: you never complete BI. It is a program, ongoing, forever changing and becoming more and more honed and sophisticated, as the business needs change. But how do you achieve this? And once you have 'done it', how do you sustain it? This presentation describes a proven BI Planning Framework , which assists organizations to achieve:

There are 7 major streams of activities which need to be simultaneously initiated, concurrently driven, co-ordinated and monitored:

STREAM 1: Corporate Data Model
STREAM 2: Corporate Knowledge Co-ordination
STREAM 3: Corporate Information Factory Development
STREAM 4: Data Profiling and Mapping
STREAM 5: Data Cleansing
STREAM 6: Infrastructure Management
STREAM 7: Data Quality Management

Many BI Programs have failed because of a lack of understanding of the real issues. The paper defines each stream and the activities inside each stream giving description, deliverable, dependencies, and duration. It also discusses team (size, skills, role), and risks and pitfalls. Attendees will learn how to establish and sustain a successful BI Program.

Derek Strauss has 25 years IT industry experience, 15 years of which were in the Information Resource Management (IRM) field. He established Data Resource Management, Architecture and IRM Functions in several large Corporations using the Zachman Framework as an architectural basis. He holds a BSc (Hons) degree from Witwatersrand University. Derek has spoken at several international conferences on IRM-related issues, including a series of seminars on BI and Data Warehousing in Eastern Europe and South Africa. He has experience in the Financial, Manufacturing, Retail, Government, Utility, and Distribution sectors of the economy and was Program Manager for a $45 million Data Quality Improvement Initiative in a large USA Bank.

A company can bring in a repository tool, and architect it into the development and production environments. However, metadata is knowledge which is most often stored in the heads of individuals who have other, more mission-critical tasks to perform. Providing dictionary definitions is not one of those tasks. This presentation shows how CIGNA has developed an approach which ensures that quality data definitions are provided to Data Administration to form a robust, completely defined metadata repository.

The REAL issue is how to get the business community to provide and validate business names, definitions, allowable values and business rules. Attendees will learn a simple approach to creating successful partnerships which will bring definitions to the table.

Stan Slossberg has 20 years in Data Administration, from dictionary/repository administrator to metadata architect. He is Chapter Founder and served as 3-term President of the Connecticut Valley DAMA Chapter; served on the Board of Officers, DAMA Boston. He is also a member of TDWI. He has designed/deployed customized repository solutions, including CIGNA's current site, supporting 42,000 employees worldwide, and getting over 8,000 hits a week. The implementation continues to grow into a more robust metadata portal. Stan has spoken at DAMA and other DA and repository sessions in Dallas, Boston, Chicago, Seattle, and other locations. Besides notable recognition in the Data Management field, he has also received the CTM certification from Toastmasters International, and has distinguished himself by receiving the Highest Level Achievement Award from Dale Carnegie.

Common Warehouse Metamodel (CWM): An Introduction to the Standard for Data Warehouse Integration

The Common Warehouse Metamodel (CWM) is a technology standard of the Object Management Group (OMG) for meta data integration in the data warehousing and business analysis environments. CWM provides the long sought-after, common language for describing meta data, based on a shared, vendor-neutral metamodel and corresponding XML-based meta data interchange facility. CWM is rapidly gaining momentum within the data warehousing and business analysis communities and is being incorporated into various vendors' next generation of data warehousing products and tools.

The objective of this presentation is to provide a comprehensive overview of the CWM standard. The primary topics covered include:

Awareness and understanding of CWM is crucial for the data warehousing and business analysis communities, because CWM is the only industry-accepted standard for meta data integration. CWM promises to greatly enhance data warehousing and supply chain return-on-investment by lowering tool integration costs and allowing for the easy combining together of best-of-breed tools, products, and applications.

John Poole is a Distinguished Software Engineer at Hyperion Solutions Corporation. He is one of several co-authors of the Common Warehouse Metamodel (CWM) within the Object Management Group (OMG) and currently leads the Java OLAP Interface (JSR-69) technology effort within the Java Community Process. He holds a B.S. degree in Applied Mathematics from Southern Connecticut State University, and an M.S. degree in Computer Science from the Polytechnic University, New York.

Dan Chang is a member of the Database Technology Institute at IBM Silicon Valley Laboratory. He has led the CWM standardization effort within the OMG since its inception in 1998 and co-chairs the OMG's CWM Revision Task Force. Dr. Chang holds a Ph.D. degree in Computational Chemistry from the University of Chicago and is a visiting professor at San Jose State University.

Doug Tolbert is a Consulting Engineer at Unisys Corporation. He is one of several co-authors of the CWM within the OMG. Dr. Tolbert is a specialist in database management systems, application development environments and related technologies, and has written and lectured on databases and meta data at industry conferences and universities for over twenty years. Dr. Tolbert holds a Ph.D. in Genetics from the University of California, Davis.

David Mellor is a Consulting Engineer at Oracle Corporation. He has been working with Multidimensional and Relational Technologies for Oracle for over 10 years. He is one of several co-authors of CWM within the OMG. He has designed and developed a variety of products specializing in meta data and multidimensional and relational technologies. Mr. Mellor currently co-chairs the OMG's CWM Revision Task Force.

"Status" is a cornerstone of business relationships and yet is rarely understood or modeled correctly. Understand and model Status to fully exploit Data Warehousing, Business Rules, and Client Relationship Management.

Attendees of this session will learn to correctly model Status; to make your database capture history accurately and derive data mart measures and facts from situations where none seem apparent. You will learn to establish a solid Status model without which it is impossible to know which business rules apply.

This is important because client and account status are natural business concepts that are seldom modeled effectively. The reasons for this are varied, but can be overcome at the start saving effort and resources. Correctly modeled status gives us new, critical information for our Decision Support environment. It ensures accuracy of classifications needed for day to day dealing with client relationships, enhancing effectiveness in both CRM and Business Rules contexts.

Henry Feinman is an information architect who has consulted with the top three banks in Canada as well as Canada’s largest City. Over 21 years of IT experience Henry has implemented operational and informational systems in manufacturing, finance and government sectors. He has worked with several data warehousing projects over the past 8 years, performing roles in data warehouse architecture, data architecture, and metadata management. Most recently, in the role of data warehouse architect, he has helped the City of Toronto implement their Enterprise Data Warehouse program and their first two datamarts.

One of the critical requirements of any Information Management (IM) group is the advertising of the unique functions that IM group performs for the company. This presentation will look at a ways to sell and to continue to re-sell the functions of IM.

Each IM groups always needs management support, but they also need to continue to help themselves by interviewing, surveying, training on the functions they perform. This presentation will provide details and examples of tools to help promote IM.

As Senior Information Management Architect for Johnson & Johnson, Larry Dziedzic is responsible for supporting global data standardization, as well as consulting on process and modeling standards for the worldwide Consumer group. This includes supporting data standardization of global ERP applications, defining data stewardship functions and interfacing with the worldwide Pharmaceutical and Medical Devices and Diagnostic groups. Larry is a Past President of Data Management Association (DAMA) in New Jersey USA, and is VP of Operations for DAMA International. Larry has presented papers at both DAMA International and DAMA US events, and also at the Enterprise Data Management Conference in Sydney, Australia. A former adjunct college instructor, Larry often guest lectures to information systems classes at the university level.

Real-life Success Story: First Steps in Consolidating Data into a Single Enterprise Architecture

Data standards are often difficult to define and even harder to implement. This presentation describes the successful data standards implementation at the MN Department of Commerce. The effort was kicked off to address inequities in the core application suite used to manage the agency’s information resources. It was defined as having an agency-wide perspective to reign in the disparate “data puddles” and application menagerie that was quickly becoming unmanageable with the goal of having a consistent applications architecture serving a smooth flowing river of data. This presentation will cover the critical success factors, noting the risks and identifying the immediate and long-term benefits.

There are many presentations of what could be done but few that cover the full breadth from concept to deployment and fewer yet that document a successful effort after it has been completed. This presentation details an unbroken chain between the identification of the need and realization of an IRM infrastructure and deployed applications.

Mark Ouska has a diverse background with a focus on methodology-driven application development and data management in both private and public sectors. He has extensive modeling experience focused on work product transformation; the science of carrying business requirements forward to implementation rather than interpreting what was discovered in the last phase.

Data Standards Implementation - And You Thought Standards Development Was Hard!

Years after developing an ISO/ANSI 11179 metadata registry, the Environmental Data Registry (EDR), to serve as the backbone to the Agency standards setting process, the Environmental Protection Agency is now tackling the challenges of standards implementation. Having successfully established a core set of Agency standards, and with new standards in progress, the Data Standards Team is conducting outreach and education activities to inform program system managers on standards implementation concepts such as stewardship, conformance, and data harmonization. Ongoing challenges include maintaining management commitment, policy development, approval processes, and standards integration. Join us for an informative talk on what we are learning along the way.

Sara Hisel McCoy is currently a member of the Data Standards Team in EPA's Office of Environmental Information (OEI), and has been actively involved with data standards outreach and implementation activities.

The Internet and World Wide Web will fundamentally change the way companies do business. Fixed relationships between product manufacturers to their distribution channels and their suppliers will be replaced by dynamic relationships managed by computer systems. These systems will conduct business on behalf of the enterprise over the Internet using real (or near real)-time message based transactions. The focus of Data Administrators will have to shift from fixed backend data stores to flexible front-end message based systems. This presentation will explain a strategy for Data Administrators so they can use their existing data models and metadata to create a model based XML internal standard and an environment to manage and control multiple XML vocabularies linked to back-end systems.

Charlie Dietz has been involved in Information Technology at MetLife for over 15 years, coming over from the Annuity Customer Service department at MetLife in 1985 to lead the development of one of the first on-line relational database systems at MetLife. Since then, Charlie has promoted data administration principles and practices throughout MetLife, most recently as creator and head of the first centralized Data Analysis and Design group in MetLife. Since 1999, Charlie has been the “XML Evangelist” at MetLife, raising awareness of XML and the need for standards to make XML an effective tool for MetLife.

The main weakness in traditional approaches to executing data migration or data integration projects is simple. They all assume that metadata is accurate. In reality, most metadata is not accurate. When you have to blow the dust off the books, you know it is out of date. The majority of traditional methods state that you start with the metadata as your source of knowledge about the system. Are you willing to bet the success of a one million dollar project on documentation that has not been updated since the system was implemented twenty years ago?

This session will teach you the basics of Data Access, Profiling and Mapping, a methodology that results in a complete and accurate understanding of 1) the proper approaches to accessing legacy data, 2) the content and structure of any data source and 3) the rules by which data must cleansed and transformed. Data Access, Profiling and Mapping creates a factual body of knowledge about your data. This knowledge will provide significant reductions in the risks and costs associated with accessing, moving, cleansing and transforming data in any type of project.

John Howe has spent 20 years working in various data management capacities. He has been a Database Analyst and Administrator, Data Modeler, Data Analyst, and Director of Data Management. Mr. Howe has been actively involved with the development and use of Data Profiling and Mapping methods and technologies for over ten years. He began his interest in Data Profiling in 1991 when he worked with the development team that created the core technology for what is now Evoke Software's Axio product line. John subsequently spent the last 7 years in various consulting, training and sales roles with Evoke Software, most recently as Vice President of Customer Services.

Data has become a critical resource to organizations. The ability to share data across lines of business allows quick responses to ever changing business needs and opportunities. Systems based on a common data architecture require effectively managed data and process models. A well designed model management strategy can provide significant savings in time and resources in the development cycle. These models are valuable assets that should be safeguarded and made available for use across the organization.

A model management strategy provides the processes and procedures required to ensure the maximum return from an organization’s data architecture investment. The strategy addresses:

The strategy should be developed and implemented in an iterative process. Each iteration of the strategy provides greater detail and greater levels of standardization. The presentation reflects approaches developed from instituting model management strategies at large retail, financial and governmental institutions.

Ralph Mohr is a consultant with Covansys. He specializes in data warehousing with an emphasis in data architecture and data quality. Mohr has been involved in a variety of projects ranging from small data marts to multi-terabyte data warehouses. Mr. Mohr's data modeling experience has been in Finance, Insurance, Retail and Welfare. Currently, Mr. Mohr is serving as the Data Architect for the Victoria's Secret and has assisted The Limited in developing corporate modeling standards and model management strategy. Mohr is a member of the American Society for Quality (ASQ), Data Management Association (DAMA), and The Data Warehousing Institute (TDWI). As a speaker, he has given data quality and data modeling presentations for local, regional and national organizations. Mohr has presented: "Determining ROI", First Logic National Conference 2001; "Architecting Data Quality in the Data Warehouse" Information & Data Quality Conference 2000 Anaheim, California; "Data Quality Assessment: A Key to Data Warehousing Success" at the Information & Data Quality Conference in 1999 in New Orleans.

One of the most significant challenges in the Data Management discipline today, is making quality estimates on the amount of resources required to do data analysis and design for existing as well as new databases. A well-respected industry standard on this is an hour per attribute. Well, there is one hitch! About 80% of the time, or more, you don’t know the number of attributes until you start to interrogate the data sources carefully. This is as you are beginning to define the logical data model.

For DBA’s, the estimates are clearly focused around the quantity of data to be processed and the complexity of the relationships between the data as defined in the data modeling process. To do an accurate job of estimating first we need to articulate what is to be included as deliverables.

    - Data model designs include:
    - Entity Identification – ie. key subject areas
    - Attribute definition
    - Attribute physical characteristics (ie. char, number, date – and length)
    - Grouping of attributes to Entities
    - Entity definition
    - Determining Keys & Dependencies
    - Row Count estimates
    - Source to Target Mapping of attributes
    - Placement of attributes into entities
    - Identifying & Mapping to necessary reports
    - Relationships between the files.

The data architect, alone in a vacuum, cannot accomplish this. It also requires the business acumen, of someone knowledgeable in the specific business vertical the data is supporting. Therefore the notion of database design is collaboration between the Data Architect, the Database Administrator and the Business Analysts and Consultants of the marketing delivery units.

Yet another variable has come into play. If a client has a base structure, but populates it with 32 different source files, can we in good faith charge them an equal number of hours to the attribute, as we would for “scratch” design?

To try to sort this out and allow us to quantify the 60 minutes consumption, I have identified those items that are purely new design separate from mapping into current structures. During our session, I plan to discuss, what I think goes into the 60 minutes per attribute estimate.

Dawn Michels is the team manager for the DW Data Architecture Team at Fair Isaac, Inc. Arden Hills office. She has 15 years experience in relational database design, including DB2, Redbrick, Oracle, NT SQL Server and SAS. Dawn has modeled many databases and guides her, as well as the refining the Data Architecture Role at Fair Isaac. To round out her career, Dawn is VP of Chapter Services of DAMA-I. She is past president and CTM of the Fair Isaac Toastmaster’s Club. Dawn is also an adjunct faculty member at The College of St. Catherine, and Univ. of St. Thomas.

This presentation will focus upon how to establish a data management presence within an organization while working at a project level. It will do so by examining a strategic vision to construct a common account management system/database for the customers, contracts and products of a single business unit within a large organization. The value of key data deliverables introduced over the course of fifteen years on both unsuccessful and successful projects will be discussed. Among them are a universal code table, data repository, universal product components and a tool for quantifying the work of data management. Attendees will learn:

Thomas Zaborsky works as a Data Analyst within the Group Benefits Strategic Business Unit at CNA Insurance, where has served for fifteen years. His responsibilities include the design and development of the SBU's databases, data models and code management. Tom holds a Bachelor of Arts degree in English Literature from the University of Illinois at Chicago and a certification from the Computer Career Program at DePaul University. Tom is active in the Chicago DAMA Chapter and is currently serving on its board in the position of Secretary.

Michael Jennings
Manager, Enterprise Performance Management
Hewitt Associates LLC

Many data warehouse projects face the decision of purchasing or building a meta data repository for their environment. The decision to build a repository is often made due to the lack of data integration between the various data warehouse products or as a result of budgetary constraints on the project. For those projects that make the decision to implement their own meta data repository solution a generic model is often the best choice to advance that effort.

This presentation explores the design of generic meta data repository data model for use in a business intelligence system project. The design will address the various base components that are typically required in meta data repository in order to support a data warehouse environment. The intent of the model is to offer those individuals who wish to implement a meta data repository a generic solution that can be integrated into an in-house implementation. Alternatively, this model design can be used to complement a requirements checklist when evaluating meta data repository products in the market place.

Michael Jennings is an architect and manager specializing in business intelligence, enterprise performance management, and web based delivery strategies & architectures at Hewitt Associates. He has more than eighteen years of information technology experience in the manufacturing, telecommunications, insurance, and human resources industries. Michael speaks frequently on business intelligence issues at major data warehousing conferences and is an instructor of information technology strategies at the University of Chicago's Graham School. He is a contributing author to the book "Building and Managing the Meta Data Repository" published by John Wiley & Sons.

Prudential has adopted the Acord XML standard to be used for all XML development. Joanne Garifo will discuss the reasons for adoption and how her organization is incorporating this standard into their metadata strategy. This presentation will discuss:

Joanne will also provide some technical information on the environment and the extensions made to the CA-Platinum Repository that allow them to capture XML and map to legacy systems.

Joanne Garifo has been a Repository Administrator for 5 years at Prudential. Responsible for all technical aspects of the Platinum Repository for MVS including database administration.

With the advent of the web, the enterprise strategy "du jour" has been focused around e-commerce. However, the web is "borderless". We tend to forget that effective exchange of data between partners and customers needs to consider differences in: Geography, Culture, Regulatory acts, etc. Visitors to our web pages and our collaborative business partners may not utilize the same content, language, or context for their data. This presentation will describe the importance of metadata for consistent data exchange and introduce XML as one possible method for describing global transactional data content. Participants will be introduced to 7 common E-Commerce and Globalization mistakes as well as the basic concepts of globalization and international data standards.

James Bean is the President and CEO of the Relational Logistics Group. He is a respected expert in the fields of Business and Technology, having completed numerous worldwide client engagements for XML Training, XML Industry Standards Development, Global E-Commerce Strategies, Information Architecture, and Database Design. He is the author of the books: the "Sybase Client/Server EXplorer" © 1996 Coriolis Group Books, "XML Globalization and Best Practices" © 2001, and has written numerous magazine articles for technology journals such as: "Enterprise Development Magazine", "XML Magazine", "DevX.com", "Web Builder CD", "PC Techniques Magazine", "Visual Developer Magazine", and the "Database Design Professional Newsletter". Mr. Bean is also a frequently requested speaker for regional, national and international technology conferences.

How long does it take to create logical, physical and dimensional data models? Your success and credibility as a data modeler not only depends on the quality of your models but also how well you estimate the level of effort up front. Setting realistic expectations is crucial! If you underestimate your time, the project is delayed and costs go up. The “sticker shock” of overestimating may cast doubts on the value of your services. This presentation shows how you can create an estimation tool that has proven its value on dozens of modeling projects at First Union National Bank.

Gary Flye is an IT Leader and Assistant Vice President at First Union National Bank, leading the Data Management function in the Database Management department. His team is responsible for database design and reengineering, data allocation and retention standards, meta data management, data quality, directory services and knowledge management. Mr. Flye holds a B.S. in Metallurgical Engineering and an M.S. in Computer Science and his 20 years of IT experience also include the mining and environmental industries in both the public and private sectors.

Centrelink is one of the largest organizations in Australia with 22,000 staff, 440 service centres and 23 call centres. It has over 5.5 million customers, makes more than 230 million payments and sends over 120 million mail items, annually. The mainframe systems cope with 12 million transactions a day. Data quality and metadata are significant issues in such an environment.

The presentation would introduce Centrelink, review the tools and processes we use to monitor and report on data quality and the important role of metadata. How data management becomes more complicated by moving to multiple service delivery channels and operating platforms, as well as new opportunities and new challenges.

Peter Davis is the Manager of Data Management in Centrelink. He has been involved with data quality and metadata issues in Centrelink since its establishment in 1997. Peter was the project manager for Centrelink's Data Quality project. An active member of the Data Management Association (DAMA) since 1996, Peter is currently the President of the ACT chapter. He has presented Centrelink case studies on the importance of metadata, data quality and enterprise data management to DAMA and other groups.

New data and object modelers are taught to hunt for all the business’s nouns as they that represent the most likely candidates for entities or object classes. Sentences with the pattern of “Noun – Verb – Noun” probably represent relationships while a generalization hierarchy often lurks behind sentences with an “is a” verb statement. Do other patterns exist in language that can help us in uncovering and structuring an organization’s business rules? This presentation strives to discover the grammar of business rules by drawing on the Zachman Enterprise Systems Architecture Framework and the sentence diagramming technique many of us learned in high school. Ms. Moriarty, president of Inastrol, an information resource management consulting firm, has enjoyed a diverse career in Information Systems, over the last 25 years, from application programmer to business analyst to information strategic planner. She has developed a methodology that integrates business rules analysis with the meta-data management environment to address major business concerns, such as Customer Relationship and Product information management. Her dynamic business models have been used as the basis of customer models for companies within the financial services, telecommunication, software/hardware technology manufacturing and retail consumer product industries.

Terry Moriarty was a columnist for Database Programming and Design Magazine for over 7 years. She currently authors the "Metaprise" column in Intelligent Enterprise. She is the co-chairperson for the Business Rules Forum, as well as chairperson for the 1997 Business Rule/Database Design Summit and the 1991 International Conference for the Entity Relationship Approach. She has been active in DAMA since 1986, where she was introduced to the data driven approach to information management and has proudly worn the badge of a data bigot ever since. She is a past president of the San Francisco DAMA chapter.

Federated Data Warehouse (FDW) Environment is used as a framework for implementation of advanced BI methods: Data Mining, OLAP and Statistical Analysis. The presentation will show how the FDW, with its conformed dimensions and unified definition of concepts (such as location, asset, customer etc.) is used to implement OLAP, Stats Analysis and Data Mining analysis in the following fields:

The presentation integrates the FDW with BI methods that are using it as a data foundation. The emphasis is put on data integration and data quality as an important element for the success of advanced BI methods and techniques. The attendee will see the big picture that will outline the BI environment including advanced BI tools applied to unified data structures integrated in the FDW.

Vladimir Pantic M.Sc., I.S.P. is Certified Senior Consultant with IBM Canada. He specializes in the domain of Corporate Data Architecture, Logical and Physical Data Modeling. As experienced practitioner, Vladimir is involved in training and education of Data Modelers and Data Architects. In last four years he is actively involved in Data Mining and Advanced Statistical analysis.

Deborah Henderson, B.Sc., M.L.S. is Data Architect for Hydro One. She is experienced in consulting to many different business functions in EIS/DSS, Knowledge Management, Data Warehousing, Enterprise Portal and On-line Analytical Processing design and Project Management, with an extensive background in information management (process modeling, data modeling and data dictionaries) and associated architecture development.

For enterprises with large data warehouses, implementing a comprehensive meta data solution can seem like a formidable task. There are no industry standards, and no off-the-shelf tool suites that can meet all of an enterprise’s meta data objectives. However, by carefully gathering requirements, mapping them to meta data sources, and choosing a solution that achieves the right balance between standardization and customization, an enterprise can develop an approach to meta data that meets its business and technical needs. Enterprises that implement successful meta data solutions will benefit from reduced development costs, user acceptance of the data warehouse, and the ability to make faster business decisions.

Thomas Gransee is a principal at Knightsbridge Solutions with more than 17 years experience in business systems planning, analysis, design, testing, and implementation of complex state-of-the-art information systems. Mr. Gransee’s experience spans the insurance, healthcare, retail and manufacturing industries. Prior to joining Knightsbridge, Mr. Gransee was a group manager in the merchandise information department at True Value Hardware. Prior to that, he was manager of retail systems at Bridgestone/Firestone Inc. Mr. Gransee has been a speaker at industry and vendor conferences, including Riscon and SCO Forum. Mr. Gransee earned his Bachelor of Arts degree in Liberal Arts from DePaul University.

A lot of what is being said, written, or done in the database management field -- or whatever is left of it -- by vendors, the trade press and "experts" is increasingly confused, irrelevant, misleading, or outright wrong. While this is, to a degree, true of computing in general, in the database field the problems are so acute that, claims to the contrary notwithstanding, knowledge, practices and technology are actually regressing!

This presentation exposes fundamentally flawed ways in which the database industry operates and illustrates some of the prevalent fallacies and their costly practical consequences. It offers an opportunity to test yourself on your ability to see through the former and avoid the latter.

Fabian Pascal has a national and international reputation as an independent technology analyst, consultant, author and lecturer specializing in data management. He was affiliated with Codd & Date and for more than 15 years held various analytical and management positions in the private and public sectors, has taught and lectured at the business and academic levels, and advised vendor and user organizations on database technology, strategy and implementation. Clients include IBM, Census Bureau, CIA, Apple, Borland, Cognos, UCSF, IRS. He is founder and editor of DATABASE DEBUNKINGS (www.dbdebunk.com), a web site dedicated to dispelling prevailing fallacies and misconceptions in the database industry, where C.J. Date is a senior contributor. He has contributed extensively to most trade publications, including Database Programming and Design, DBMS, DataBased Advisor, Byte, Infoworld and Computerworld and is author of the contrarian column Against the Grain.

UML data modeling profile gives database developers the opportunity to use object oriented design methods for the design of a data warehouse. Davor will explain the Unified Modeling Language in context of database modeling.

The session will introduce the design of a star schema and snowflake using object oriented design and UML. Different diagrams and the semantics of UML will be introduced using a design example for a data warehouse. Themes of the session include:

Davor Gornik: After the study of computer science in Munich, Davor worked on software projects for database applications in retail and finance area for different companies as developer, consultant, and project manager. He joined Rational in 1997 and worked as a technical representative. He joined the Rose Business Unit as a marketing engineer for the Data Modeling product.

The Politics of Data Analysis: Working Within the Constraints of Corporate Data

Information Quality is the latest buzzword of the 21st century - and rightly so. The consequences of poor information can be severe enough to impact more than just business operations. An extreme example is the recent merger of two major banking institutions. During the merging of the consumer information, funds in one customer account were electronically lost causing severe personal trauma. Many corporations are recognizing the negative impact of poor information quality and are investing in improvement. This should create jubilation since historically, data analysis, has been perceived as low value, especially on time-constrained projects. Excitement wears off quickly, though, as it is apparent that in large corporations, poor information quality has roots so deep, that in spite of best intentions, most barriers remain in tact. Ironically, even corporate solutions intended to improve Information Quality, such as the Enterprise Data Warehouse or Enterprise Applications also serve to broaden the spectrum between good data management and reality. Regardless of the root cause of poor data management practices, the need to appropriately apply theory and practicality is critical to successful data analysis.

This presentation intends to assist data analysts who are required to work within the corporate bounds of poor data management practices. Using a case study from a Fortune 500 manufacturing company, common constraints encountered in corporate environments are identified and best known methods that mitigate the impact and propagation of these practices are examined. The presentation is not intended to advocate eliminating the use of Information Quality theory, but strives to create recognition of the conflict between theoretical practices and practical application in large corporate environments.

Amanda McLoone has over 13 years experience in the Information Systems industry with an emphasis on Data and Business process analysis. She has been employed at Intel Corporation for the last 9 years in a broad range of organizations covering IT, Logistics, Factory Automation, Corporate Quality and Supplier E-Business. She currently manages the Corporate Quality Business Process Engineering and Data Analysis groups. Amanda’s career accomplishments include the original implementation and proliferation of Intel’s corporate shared data environment. She is responsible for achieving significant cost savings through the automation of a major business process. This cradle-to-grave inception of automating Intel’s Capital Equipment Forecasting process and its delivery as Intel’s first business-to-business application was a major breakthrough for Intel’s eBusiness initiatives. Her most recent accomplishment is the adoption of Business Process Quality as a core function within the Corporate Quality Network.

The corporate world is drowning in disparate data. Data elements, a.k.a. field names, column names, row names, labels, metatags, etc. seem to reproduce at whim. Librarians have been battling data disparity for over a century with tools like controlled vocabularies and classification schemes. Data Administrators have been waging their own war using data dictionaries and standards. Both camps have had limited success with their strategies. Why? (1) These tools are not a common data architecture and in some ways actually contribute to the problem. (2) Librarians and Data Administrators are not working together to find a total solution.

Sandy Hostetter, a professional librarian, is currently employed by the Rohm and Haas Company as manager of the Content Management Group within the company’s Knowledge Center. The Content Management Group encompasses the areas of records management, document management, and metadata management. Sandy has spent most of her career in library-related positions at Rohm and Haas, but a few years ago she went over to the “dark side,” and labored for almost 2 years as the metadata architect for a large data warehouse project. During this period she was fortunate to “discover” the writings of Mike Brackett, and since then has been actively collaborating with him on a creating a common data architecture for Rohm and Haas.

Today’s packaged software includes rich meta data repositories, which are used for the configuration and operation of the software. The package meta data is useful beyond the primary system, especially for those systems that share data. However, the package meta data frameworks often provide levels of abstraction that obscure the underlying semantics of the system and its data. Some shared data will likely be implemented in incompatible ways in the different systems. This presentation relates the experience of a real-word project that integrated the meta data repositories of several packaged software products, including CRM, Billing, ETL, and EAI toolsets.

With the rise in packaged software and decline of custom development, meta data has been moving from custom-developed models to vendor-developed models. As always, it’s important that system models be managed appropriately. In a way this is no different from the meta data management of custom development. However, the packaged repositories provide some new challenges, which this presentation addresses.

Van Scott is a results-oriented data architect with a very strong methodological bent. He is intrigued by the enterprise, enterprise architectures, and data warehousing. Very project-focused. Published author. Speaks at business intelligence seminars, including the DAMA 2001 conference.

The Corporate Metadata Repository (CMR), is an Oracle-based, commercial, off-the-shelf (COTS) software package that stores metadata about the Bureau of Land Managements (BLM) applications. By creating a central, shared source of metadata, with the consistent data definitions, CMR reduces application development and maintenance costs. The CMR is BLM's data management tool used by data administrators, database administrators, developers and other application users. It is the BLM's reference for standard data elements and metadata.

Theresa Fresquez has worked for the U.S. Department of the Interior's Bureau of Land Management (BLM) for the past 10 years. Throughout her years at BLM, Theresa has worked primarily on database development projects and has developed several wildlife-related national applications. Most recently, she has been involved with developing and populating BLM's Corporate Metadata Repository. Theresa has a B.S. degree in Computer Science from Colorado Christian University. She has received several awards and recognition for her achievements at BLM. Theresa is a member of the DAMA and Platinum Repository User Group.

Brian Timothy Campbell - Over 25 years experience with US government DOD computer systems including 4 years experience with Bureau of Land Management systems. Mainly a hands-on engineer that has maintained a technical career path, and has developed extensive skills in small Rapid Prototype Development efforts and large computer system lifecycle projects to include requirements analysis, system engineering, software development, database modeling and maintenance, integration and test, and project management.

Part 1:How to Organize Data and Content Across Multiple Applications and Databases

Meta data is an integral component in organizing content seamlessly across multiple databases and applications. With the proliferation of taxonomies, thesauri, keywords, and other organizing schema, how can you best navigate through, evaluate, and utilize the meta data that is most useful to you? This session will address the steps needed to leverage legacy meta data and also create new meta data when none is available for classification of data within a single, optimized system. Some of these steps include:

Lynn Wojcik is the Director of Content Classification and has worked at Northern Light Technology since 1997. She has been the Chief taxonomist at Northern Light since 1999. Before coming to Northern Light, Lynn held several positions at the Office of Smithsonian Institution Archives and the National Digital Library Project at the Library of Congress. She has a BA from Smith College and a MLS from The Catholic University of America.

Significant effort is spent by enterprises to reconcile the messages and transactions exchanged between computer-based systems. The key issue rests with contextual gaps that arise when different systems (eg. front office vs. back office) transmit and receive data about the same concept using their unique vocabularies and implied meanings. Context gaps represent a significant (and hidden) recurring cost in enterprise application integration (EAI) and business-to-business (B2B) commerce. This presentation discusses existing and emerging solutions to the data context problem using case studies from the financial services and environmental science domains.

The speaker will describe real-world deployment problems and their resolution; the successes and the failures; and the “soft spots” to monitor in related implementations. The presentation concludes with a practical perspective on emerging initiatives that address the context gap problem (eg., “Semantic Web”, RDF, DAML, Agent-Based Information Brokers, and data standardization efforts).

This presentation will engage attendees with current views of the “state of the art” in data management. More importantly, it will describe in plain English the significance and potential impact of various technologies and data standardization initiatives.

Chito Jovellanos is the President & CEO of forward look, inc., a Boston-based company that helps its clients secure recurring new revenues from their enterprise data assets. Prior to forward look, he served as the Chief Information Officer of Internet Securities, Inc (a Euromoney company). Chito was also Director of Research & Development at the Electronic Settlements Group (ESG) at Thomson Financial Services; Vice-President of Business Planning at Scotiabank; and Director of Product Development in the Transaction Products Group at Reuters plc. He is a professional member of the Association for Computing Machinery (ACM), the Data Management Association (DAMA), and the International Association of Financial Engineers (IAFE).

This presentation is focused on the "How" to develop an Enterprise Data Model. It describes the approach developed and used at Delta Air Lines for the creation of an Enterprise Data Model. The Delta Air Lines Enterprise Data Model is now being used to create the Operational or Enterprise Data Stores, integrating operational data across the airline business. It describes a 7 step practical methodology for developing an Enterprise Data Model that incorporates a "top Down" and "Bottom up" approach. It incorporates an enterprise view needed for integration to support an ODS and/or DW, as well as the current state (work already accomplished – existing models) for practicality and quicker development. The presentation focuses on How to build the enterprise data model using this methodology.

This is not an academic or theoretical presentation. It is taken from a real corporate successful experience of actually producing an Enterprise Data model. The model is now over 700 integrated entities and encompasses 5 of the major business subject areas. The model is being used to build the Operational Enterprise Data Stores and will eventually be used for the Enterprise Data Warehouse.

Noreen Kendle - I have been with Delta Technology (the IT of Delta Air Lines) for over 6 years, presently as an Enterprise Architect. When I developed the Enterprise Data Modeling methodology I was the Manager of the Enterprise Modeling group which composed of 24 modelers. We have been working on the Enterprise Data Model for nearly 3 years. We have developed extensive modeling, data, naming, XML, review and integration standards. Prior to managing the enterprise modeling group I worked as a data architect across the airline on a variety of projects. I have been involved in the development of the ATA airline model. I have worked in and around Data for much of my over 20 year career. Prior to joining Delta I worked for AT&T as an Oracle DBA, Unix Admin, Data Modeler and application developer. Prior to AT&T I worked at Masco Corporation, EDS - World Computer, AT&T Corporate/Chrysler, and Ford.

C. Lwanga Yonke
Manager, Data Architecture & Information Quality
Aera Energy LLC

When properly designed and implemented, an Information Stewardship program can successfully create the right accountabilities for Information Quality. This presentation describes Aera Energy’s approach to building and nurturing an information stewardship culture. A deliberate effort was made to move beyond data definition to include stewardship accountabilities throughout the information value chain (create, enter, update, apply, delete). Topics include:

C. Lwanga Yonke works as Manager of Data Architecture and Information Quality for Aera Energy LLC. He is actively involved in the implementation of an Enterprise Architecture Plan and of a dataware methodology for data warehousing and system development. Lwanga currently specializes in all aspects of information quality. In previous assignments, he led multiple development and operations projects in petroleum engineering. Lwanga earned an MBA Beta Gamma Sigma from California State University, and holds a bachelor of science in petroleum engineering from the University of California at Berkeley. An ASQ Certified Quality Engineer, he has conducted numerous workshops and seminars on TQM strategies, and has authored and presented several technical and IRM papers.

Discussion focuses the value of the Data Analyst within OAOD; Issues with the Object/Relational approach (specifically using UML-Class models to represent data); What the industry is doing from a tool perspective; Data Analyst role in Object Oriented Analysis, Design, and Testing as well as the benefits of data modeling throughout this process.

Attendees should walk away with a clear understanding that with the OAOD approach, the Data Analyst helps:

Jim Goetsch is the Data Architect at Schneider Logictics, Inc. and an Adjunct Professor at St. Norbert's College in Green Bay, Wisconsin. Jim has been working in a variety of roles within Software Development using relational databases since the mid-80's. He has a B.S. in Computer Science and Engineering from the Milwaukee School of Engineering and a Masters in Business Administration from Cardinal Stritch College.

The Metadata Services Group within BellSouth has spent the last 2 years developing Metadata based products. However, the successful collection of metadata is only half the battle. Organizations that develop metadata solutions must begin to transform themselves into a services based organization. Failure to focus on services will result in another failed metadata project. This presentation will review the basic steps of transforming the metadata group into a Metadata Services Group.

R. Todd Stephens is the Director of the Metadata Services for BellSouth in Atlanta, GA. Todd has been with BellSouth for about 4 years. His primary responsible is setting the corporate strategy and architecture for the development and implementation of Metadata Repositories, which include metadata, data transformation, component, XML, content, documentation, metrics, interfaces, and the Enterprise Information Portal using the XML technologies. Todd has developed frameworks that have earned him two patent-pending applications within the past 2 years. Todd is enrolled at Nova Southeastern University pursuing his Ph.D. in Information Systems. Todd’s area of research interest include Metadata Reuse.

Brett Champlin, Allstate Insurance
Karen Lopez, InfoAdvisors
Robert Seiner, TDAN.com and CIBER, Inc.
Graeme Simsion, Melbourne University

This lively and entertaining panel session will tackle all of the latest issues and technologies, as we attempt to sort between the stuff that you should be paying serious attention to, and buzzwords you can avoid wasting time on. As the conference draws nearer we will define the topics more specifically (to make sure we cover the really "hot" topics), so send us your ideas, and check on the conference web site for session updates. Just to get the ball rolling, here's a few initial issues:

- Wireless data
- Real-time data warehouse
- Zero-latency organizations
- Web services
- Privacy
- Data management outsourcing
- Message brokers

According to Information Week, 89 percent of companies are taking on Customer Relationship Management projects. All of these projects involve data management, but few of them plan for it. Many of the pervasive press reports heralding the spectacular failure rates of corporate CRM programs have as much to do with failures of implementation as they do with failures of strategy. In this presentation, author and consultant Jill Dyche will review what data administrators, DBAs, project managers, and business analysts need to know about preparing for and implementing a CRM project--the first time out. Her talk will include several real-life case studies from companies who've succeeded and failed at CRM.

Jill Dyche is Vice President of Management Consulting for Baseline Consulting Group, a firm specializing in the design and implementation of customer databases for businesses from start-up companies to the Fortune 50. She is the author of e-Data: Turning Data Into Information With Data Warehousing, which has been published in four languages. Ms. Dyche consults to both Baseline clients and to the vendor community on planning for customer-focused technologies. Ms. Dyche’s writings have been featured in Intelligent Enterprise, CIO Magazine, Information Week, Computerworld, and The Chicago Tribune, among others, and has lectured globally on the topic of managing data as a corporate asset. Her new book, The CRM Handbook, clears the haze around the different types of CRM initiatives and how companies should plan, design, and deploy them, and is a business best-seller.

After twenty-five years in development, data models (and their cousins, Object Models) appear to have come into their own in our industry. The problem is that there are many different approaches and many different attitudes toward data modeling. It is becoming clear, however, that some characteristics of data models are better than others. This paper is one man's attempt to articulate just what those characteristics might be.

The first premise of this presentation is that the quality of a data model is directly proportional to its ability to communicate concepts to non-data modelers. The data model is a communication tool to establish that an analyst's understanding of the nature of an enterprise is in fact correct. If it cannot do that, it is not worth the effort. Data models can be evaluated in terms of their graphics and the way they are presented. This paper will discuss both, including issues of the notation to use, care in creating models, and the way they should be organized.

This presentation describes the characteristics that should be part of any data model, if it is to be successful in supporting a requirements analysis effort. It addresses the question of how to make a data model readable to someone who normally doesn't read data models. (And yes, it is full of biases and prejudices - but they are held by someone who has been a very successful data modeler for over fifteen years.)

A veteran of the Information Industry since the days of punched cards, paper tape, and teletype machines, Dave Hay has been producing data models to support strategic information planning and requirements planning for over thirteen years. He has worked in a variety of industries, including, among others, power generation, clinical pharmaceutical research, oil refining, forestry, and broadcast. He is President of Essential Strategies, Inc., a consulting firm dedicated to helping clients define corporate information architecture, identify requirements, and plan strategies for the implementation of new systems. He is the author of the book, Data Model Patterns: Conventions of Thought, published by Dorset House, and, more recently, producer of Data Model Patterns: Data Architecture in a Box™, an Oracle Designer repository containing his model templates.

What’s So Spatial About Spatial Data? Integrating Geospatial Data into the Enterprise Data Resource

Geographic Information Systems (GIS) technology evolved separately from other information management technologies. This was fine when GIS concentrated on computer cartography and left management of non-graphic attributes to mainstream DBMS technologies. Now, however, improved GIS and RDBMS data handling technologies and increasingly urgent business needs are forcing a convergence in these two aspects of data management -- to the consternation both of GIS staff who don’t understand the complexities of general data administration and of IT data administrators who don’t understand geographical data.

The fundamental challenge is that geospatial data within a GIS adds whole new layers of complexity in addition to those data administrators are accustomed to dealing with when incorporating more traditional data types within the DBMS. Even the inclusion of imaged documents or raster photography could be accommodated relatively painlessly, once we had data types like BLOBs to handle their unique storage needs and raster viewing tools available for their display. In contrast, GIS features have geometry and explicit locational coordinates expressed in specialized and quite technical units of measure. They often have topological characteristics and relationships that further extend the bounds of data management.

Bringing the power of the “intelligent map” that is GIS into our business operations is clearly of great benefit to our organizations. Fortunately, despite the “spatial ness” of this data, it can readily be accommodated in our data management activities once the few basic concepts presented during this presentation are grasped.

Mike Walls is Software Engineering Manager for PlanGraphics, Inc., a world leading consulting firm specializing in GIS and its integration into enterprise information technology and management. He specializes in project management and data architecture issues, but still works on complex data modeling and database design challenges as needed. Prior to joining PlanGraphics, he worked for over 20 years in local government as a policy analyst, city planner, applications programmer and systems administrator.Mr. Walls has academic degrees in anthropology, public administration, and computer science. He has also recently completed 25 hours towards a doctorate in geography. He has over 16 years of hands-on experience designing and implementing GIS. In 1999, URISA published his “Data Modeling” in their Quick Start monograph series.

There are many different approaches or architectures for developing data warehouses, such as centralized, functional, federated and virtual data warehouses. For data marts, there are imbedded, dependent and independent data marts. Each one has its pros and cons, and each its special considerations. But they are not all equal. Some can deliver short-term value but over time will increase total cost of ownership. Others will flatly leave you dead-ended. Some will create more work than they are worth over the long haul. Others are difficult to do, even though it may be the right thing for a given customer. It is important to balance short-term gains with long-term benefits. The choice of correct architecture depends on such factors as: goals of the business, maturity of the organization, data and queries, centralization of the organization, and commitment of the business sponsor. This choice is complicated by a lot of bad or self-serving advice in current data warehouse literature. The age of eCommerce empowers it more easily to deliver value to the business. This presentation will present case examples of successful and unsuccessful warehouses. The most powerful current trend will also be discussed. The answer may surprise you.

Thomas Haughey is one of four originators of Information Engineering in America. He is currently CTO for Pepsi Bottling Group after being Pepsico’s Director of Enterprise Data Warehousing. He was formerly President of InfoModel, Inc., a consultancy in Data Warehousing. His courses have been delivered to companies around the world. He has worked on the development of seven different CASE and wrote his own in 1984. He formerly worked for IBM for 17 years. He is the author of many articles on DW and IE. He was VP of Technology for Silverrun Technologies. He is working on a book, "Designing the Data Warehouse - The Real Deal". Tom earned a BA in English.

This presentation examines the major commercially available products for integrating and managing meta data from heterogeneous sources. Major features and capabilities of each tool are reviewed and compared, with their relative strengths and weaknesses.

Attila Finta and Marcia Rhode are the founders and principal consultants of A.M. Consulting, a business and IT consultancy based in Austin, Texas, specializing in data warehousing and business intelligence, systems planning, and meta data management, with experience in manufacturing, retail, airlines, utilities, telecommunications, finance, insurance, petroleum, pharmaceuticals, and other industries. Mr. Finta and Ms. Rhode have been active in the data management community for many years, are past board members of DAMA International, and are co-founders of the Heart of Texas DAMA Chapter.

A Case Study on Using XML to Collect Data from Multiple Sources and Render Multi-Media Reports

This presentation will describe an internal Information Technology project to create multiple financial and status reports from various, diverse data sources. The data needed for the reports resides in various depositories, including Oracle, proprietary and text databases. The desired output formats include word processing documents, spreadsheets and web pages. Both internal reports, and edited versions of the various reports for external publication are required, so the issues of confidentiality and proprietary information need to be considered and resolved. Weekly periodic reports with multiple sorts and filters are required, as well as monthly and year-to-date summaries and roll-ups.

The attendees will see how XML is truly platform and database independent, and can be used to transform data from virtually any format into any other format, including web pages and forms. Major points to be covered in the presentation include:

Jim Grosso serves as a Senior Project Manager specializing in Electronic Commerce at Kanbay, Incorporated, a global information technology consulting firm. He has more than 25 years of diversified experience in information technology and manufacturing engineering, with over 20 years in Electronic Data Interchange and Electronic Commerce, and has received certification from IBM in “XML and Related Technologies”. Mr. Grosso has given presentations on EC, EDI, Automatic Identification and other technologies at many conferences and seminars, and has written articles for various publications including ActionLINE, the EDI Forum, Logistics Information Management and Fabricator Magazine. Jim is a two-time winner of the Outstanding Achievement Award from the Automotive Industry Action Group, and has received professional certification from the Project Management Institute.

Whether or not object storage is new to you, it's been around for quite a while. Object storage is now becoming increasingly important as people are looking at Java application servers and other ways for taking advantage of the Internet. Looking at it beginnings, current state, and what the future holds will provide a broader view of object storage and what it can do for you. Join Doug Barry as he mines his more than 14 years of work in object storage to provide a provocative look at where the industry has been and where it is going.

Doug Barry has worked in database technology for over twenty years, with an exclusive focus on the application of database technology for objects since 1987. As principal of Barry & Associates, Doug has focused on helping clients make fully informed decisions about the application of object technology. He emphasizes the need to match product feature strengths to application needs in order to ultimately field a successful product. Doug is also the author of the Object Storage Fact Book for Object DBMS and Object-Relational Mapping products, published by Barry & Associates, Inc.; the Object Database Handbook: How to Select, Implement, and Use Object-Oriented Databases, published by John Wiley & Sons; the XML Data Servers: An Infrastructure for Effectively Using XML in Electronic Commerce, published by Barry & Associates, Inc.; and was for many years the Databases columnist in Object Magazine and the ODBMS columnist in Distributed Computing Magazine. His articles have also appeared in Database Programming & Design, Intelligent Enterprise, IEEE Computer, Software Development, Component Strategies, and Data Management Review. In addition, Doug serves as the Chair of the Object Data Management Group (ODMG), a consortium of vendors and interested parties working on object storage standards.

Davida Berger, VP Conferences, DAMA International, and
Enterprise Meta Data Architect, A&P

Entity Relationship Modeling (ER), Object Role Modeling (ORM) and Unified Modeling Language (UML) are techniques used for data and object modeling. There has been much controversy among practitioners regarding the appropriate use of each technique. In this lively, interactive session, our panel of international experts will provide an overview of each modeling methodology and discuss the pros and cons of each technique.

Davida Berger has fifteen years experience in information technology. Her experience includes management roles for data centric projects including an enterprise data warehouse and the start up of a successful data administration function. Davida has written columns for DAMA, and was the coordinator and moderator of The Data Administration Newsletter (www.tdan.com) interactive Reader's Corner which brought together industry experts in an on-line question and answer forum. She is the President of DAMA NJ, and has been the VP Conference Services for DAMA International since 1999. She is a nominee for the 2002 DAMA International Individual Achievement Award.

Using a Data Architecture Approach for Getting from Strategy to Implementation

In order to be prepared for a fast-changing business environment, the information resource must be constructed according to a well-considered data architecture. By adding transformation rules to the Zachman Framework and following a solid methodology, an IT organization can implement an information resource that embodies the goals, mission, and vision of the business people. This presentation will show how to start with models that reflect the information needs of business people, how to transform those models within the context of the architecture framework, and to build models that implement the business vision. The presenter will use examples and will discuss the transformation rules necessary for moving through the Zachman Framework. Another facet of the discussion will be how to build and organize a data architecture effort that will ensure ongoing synchronization of the data resource with the business.

Michael K. Miller is the Manager of Data Management in the Cardinal Distribution IT organization. His background is in general business management, computer systems development, systems analysis, and IT architecture. He has practical experience with all phases of architecture, methodologies, tools, process management, and data management. In his career, he has had important roles including chief architect, chief data architect, technology architect, methodologist, senior consultant, and development manager.Mr. Miller has been a frequent presenter at DAMA Symposia and other industry conferences. He also authored articles for the DRM Journal, the DM Review, and several editions of Handbook of Data Management. He was the first President of the Southwest Ohio DAMA Chapter.

Danette McGilvray
Agilent Information Quality Program Manager
Agilent Technologies

Many companies are migrating data from disparate systems to integrated applications or creating warehouses to consolidate information. Why is this an opportunity to improve information quality? How can you integrate information quality into any data migration initiative? What do you need to do to ensure the long-term health of the information on which your business relies? What preventive/proactive measures should be put in place? Learn answers to these questions and find out how Agilent Technologies is implementing these important principles.

Danette McGilvray is information quality program manager for Agilent Technologies. She is implementing a program for managing information quality on an enterprise-wide basis and is currently working on a large-scale data migration project at Agilent. She consults with and leads project teams and speaks frequently at industry conferences. She was featured in a PC Week article about data quality and some of her program's best practices are highlighted in Larry English's book Improving Data Warehouse and Business Information Quality. Her experience includes information resource management, direct marketing, and electronic data interchange.

This presentation focuses on the ANSI/ISO Standard, 11179 for Data Element Metadata. What does it meant to conform to 11179? The presentation materials were created to answer that question.

The materials provide the critical definition for a data element, and proceeds through the key sections of Part 3 of the draft 11179 standard illustrating the different metadata models that must exist to then claim conformance to 11179. To make 11179 "real," the materials bring all the 11179 high-level models together into one "relational" data model for data element metadata. Finally, the materials illustrate the necessary "down-stream" metadata models required to implement 11179 data element metadata models within a metadata repository that supports enterprise-wide data standardization. The benefits of the repository are then illustrated through an examination of several collections of metadata from "real-world" database applications.

Michael M. Gorman has been involved in data processing, database and DBMS for 35 years. He has been secretary of the ANSI NCITS H2 Technical Committee on Database for over 22 years. H2 participated in the development of the 1986, 1989, 1992, and 1999 SQL standards.In addition to 22+ years of ANSI Standards development, Mr. Gorman worked for the System Development Corporation which in conjunction with MITRE and Lincoln Labs invented data management. Mike was the Federal manager for MRI Systems Corporation, who's data management system, System 2000, was deployed throughout the Army, Navy, Air Force, EPA, HUD, Commerce, USGS, the USDA, and a number of other agencies. He was the Federal Database Product Manager for the Computer Sciences Corporation. Mike authored Yourdon's database project and database administration courses. He also was a member of The MITRE Corporation's technical staff and the conference and program manager for MITRE's Data Management conference. Mr. Gorman has presented to DAMA 2000, DAMA 2001, WRAD, and DAMA chapters in New Jersey, Chicago, Philadelphia, Sacramento, St. Louis, Minneapolis, and Washington, D.C, and also to SHARE/GUIDE. Mike has also taught for several years at the Washington, D.C. ACM professional seminars.Mr. Gorman has published articles for IDC, Auerbach, Computer World and several ACM journals. He has published database and DBMS books through Q.E.D. Information Sciences and John Wiley and Sons.

Some companies have a difficult time justifying a large expenditure to purchase commercial meta data products. This practical presentation will show attendees that useful solutions can be built with the products you already have and a little bit of hard work. Areas covered will include logical meta data and model management, physical meta data, environment meta data and linkages to change management and CASE tools. An effort will also be made to align with the Zachman Framework.

Jeff Davey - Data Architect with 15 years experience. Worked in several industries including Manufacturing, Retail, Distribution and High Technology. Worked on large OLTP and Data Warehouse designs. Also spent time working with SAP R/3 as a project manager. A member of DAMA International via the Chicago chapter since 1993 where he has served on the board as VP Programs, VP Online Services and is currently their President.

ERP systems change the nature of data warehousing in a company. Some usual difficult problems in the data warehouse are greatly simplified, but other aspects are much harder. The use of abstract representations of data in the ERP compared to the concrete representations needed in the data warehouse requires an understanding of both levels and how to map the two. We will cover the simplified parts and how to deal with the more complicated parts of the process.

Mark Riggle is founder and CEO of Causal Aspects, a business intelligence software company. Mark has been in software research and development for more than 20 years. His background covers research in Artificial Intelligence, programming language design and data warehousing, most recently at SAS Institute. He has a BSEE and graduate degree in computer science from Rice University.

This session describes a new information modeling technique and associated toolset which addresses the fact that reviews of an information model by business stakeholders, subject matter experts, etc. often fail to identify flaws in the model. This is because models are most often presented in a diagrammatic manner which is fine for a high-level overview but deficient in two respects:

By contrast, the technique described in this session generates a set of assertions about the model which can be reviewed individually, each reviewer disagreeing, agreeing or seeking clarification. Once an agreed assertion set has been achieved (by changing the underlying model), the model can then be used to generate a database design. Information modeling has two audiences: business stakeholders and database designers. Diagrammatic techniques do not adequately serve to assure effective reviews of information models, so that many systems are built or acquired which do not adequately meet business needs.

Graham Witt - In Graham's 30+ years of industry experience, he has developed specialist expertise in system specification, user interface design, data modeling, relational database design, data quality, the use of metadata repositories and CASE tools, and information management. As well as completing a number of successful projects in these areas for major clients in such industry sectors as education, health, telecommunications, finance, transport and government, he has a reputation as an effective educator in these disciplines, and has published in the information management press.He is the co-author, with Graeme Simsion, of the second edition of the seminal text “Data Modeling Essentials”.

Every organization has information that it needs for survival. Examples might be a list of clients and receivables. But, organizations collect and use more information than just that needed for company survival; e.g. some may be extra client information that helps fine-tune the business relationship.

There are degrees of how important information is to the organization, and it can be compartmentalized for better understanding. This paper will show the categories of corporate information now being used in the BC Ministry of Forests. It will also describe language that can be used to promote information management in an organization.

Jeremey Janzen, a designated Information Systems Professional of Canada (I.S.P.), has been actively working in the systems field for twenty-one years, the past sixteen concentrating on developing a data administration function and promoting information management within the British Columbia Forest Service (BC Ministry of Forests) in Canada. He has chaired the BC government-wide Data Administration Forum since its inception in 1996. Jeremey feels strongly about the importance of managing information as a corporate business resource, and that it is necessary to re-build how people think about information (a.k.a. "culture change") for organizations to be successful. Jeremey brings a practical rather than academic approach to spatial and attribute data management, with a focus on realizing short- and long-term gains and "implementing goodness"!

The question for many is not related to the need for and importance of Enterprise Architectures. Rather, the question is: How do we (or can we) succeed in creating an architecture and ensure that it is used?

This presentation provides multiple real-world results of successful architecture efforts. It demonstrates two approaches—a data driven approach to create the knowledge base foundation of an architecture and a product-based approach to distribute and jump start its use. It also addresses a process to maintain the architecture through the synergistic integration of data and IT management processes.

No matter how much we recognize the need for and importance of architectures, the bottom line is, an architecture has to be produced and it has to be used to be of value to an enterprise. This presentation helps address the questions:

Richard Hecht is President of DATA Architects Technicians Analysts, Inc., a consulting firm providing Enterprise Architecture, IT Planning, and Enterprise Modeling services. As a corporate IT employee, Price Waterhouse senior manager, and founder of DATA in 1992, Rich has 25 years of IS experience with over 30 companies in 16 different industries including financial, government, defense, healthcare, manufacturing, insurance, and transportation. Recent engagements have focused on creating and integrating enterprise architectures into IT planning, management, and development processes. Rich has presented at several conferences, taught classes, lectured at universities, and served on an international committee for developing data modeling methods.

In an uncertain future, the only certainty is that things will not be the same as they are now. There are very few who predicted the phenomenal changes in the stock market and individual company’s fortunes. The horizons for strategic planning have been significantly reduced from years often to weeks. In this environment of instability, management practices have run the course from re-engineering the enterprise to optimize business processes to a notion of “react and adapt”. The price that an enterprise pays for a lack of nimbleness is loss of market share, customers and eventually profit.

In the past several years we have seen a number of new approaches appear in the knowledge and information management arena to deal with the speed of change. Evidence can be seen in the number of trade shows and conferences that relate to ERP, CRM and the ubiquitous e-ks (kitchen sink). There is a new area of concern that has arisen as a result of the many failed attempts to make IT more responsive to the changing requirements of the business. It is within the Enterprise Application Integration (EAI) that we see the results of these many attempts. Most companies shave a significant investment in an ever more complicated wiring diagram that maintains the interfaces between the niche solutions and the enterprise wide information infrastructure.

In the information architecture world, data is the structure around which we build and use applications. In fact it is actually meta data that forms the underpinnings of information and knowledge management.

The session will explore how the effective use of meta data can help enterprises deal with change. The impact of change will be covered from the perspectives of change from within and change from without, whether that change is planned or unanticipated. The path for evaluating the impact of change will be traced from the business and technical domains (i.e.: new business merger strategy, integrating new technology).

There will also be an overview of the tools and techniques that will assist the knowledge management professional in deploying a successful meta data implementation.

Don Soulsby is Director, Industry Liaison for Computer Associates Information Management Marketing and is based in Toronto, Canada. His areas of specialization are strategic and technical architectures for metadata management, data warehousing and business intelligence. He has an extensive background in systems architecture and planning, business requirements definition and data modeling methodologies with over 20 years of experience in the development of operations and decision support applications.

Is it possible that XML enables a more faithful “version” of Relational Theory than the current generation of DBMSs and SQL? Mark Milodragovich believes so, and will present several specific arguments for how XML, XML query languages, and XSL can offer advantages over current implementations of relational databases and query languages.

As we invent the Semantic Web we have a new opportunity to get it right, or at least get it "righter".

Mark Milodragovich is a Senior Information Engineer at Nimble Technology, an XML-based data Integration company based in Seattle, Washington; Fourteen year veteran of The Boeing Company mentored by Data Resource Management expert John Hancock; Developer and manager of many enterprise level applications and databases.

The presentation will draw on practical experience within the British Army to describe a new modeling language that simplifies the problems of data standardization. This language has been developed by the Army to describe information rather than data. It is used to define different types of real world object together with the types of information by which they are described. The resulting models are wholly independent of system implementation. The presentation will show how such models can be used as a standard that can be endorsed by user communities and mapped simply onto any number of different data designs.

Harry Ellis is senior consultant at Army Data Services in the British Army. In this role during the last eight years he has provided the technical lead for the development of both CBML and the underlying quantum datastore. Harry was one of the early pioneers of entity modeling in the late seventies and was the main inspiration behind what has come to be known as “the Barker Method”. He has a wealth of practical experience having worked at all levels from chief programmer to corporate information systems manager at Smiths Industries. At International Computers Ltd (Fujitsu) he was responsible for the technical specification of all data management products including IDMS-X and the first data dictionary to provide support for information modeling. As an independent consultant since 1982 he has created corporate data models for many major enterprises such as Ford of Europe and Mobil. As advisor to the British government he wrote their handbook on corporate data modeling and compiled meta-models of many system development methods and tools.

To manage, consolidate and cross reference data content from different data suppliers in different languages and at the same time manage knowledge, quality and processes has become the key to survival for anyone in the data management industry.

Data model, metadata and knowledge management considering both business and operational needs are part of a data strategy that has to pack local needs into global views... and yet maintain cost effective and competitive data operations. In the presentation Stephan will cover:

Stephan Stadelmann has over 15 years of experience in the financial information industry in Asia, USA and Europe. Providing advice, assistance and designs data strategies for data vendors, exchanges and financial institutions. Covering data management/architecture /warehouse, metadata, migration and streamlining in areas like market data & securities processing, back office, asset management, reporting, research, standards and data productization.

Find out how managing essential pieces of data, including the allowable codes and their meanings, is the first step toward supporting business rules consistently. Who are the key players in this process? How do you make the leap from managing data to providing it to applications for use in enforcing data constraints? And when your primary business client decides a package is the answer to all their problems, how do you integrate existing data (and rules) into that packaged environment with a minimum of pain? Barb Morgan and Cathy Nolan will describe the approach taken at Allstate, and demonstrate the tool that manages the coded information and controls the deployment of this data to the applications.

Barb Morgan is a Team Leader responsible for Repository and Meta-Data Management at Allstate Insurance Company. She has 28 years of experience in Information Systems.

Cathy Nolan is a Codes Analyst for Allstate Insurance with 18 years of experience. Both Cathy and Barb are Past-Presidents of DAMA Chicago and severed on many other Board positions. They are co-editors of "Guidelines to Implementing Data Resource Management" published by DAMA-I.

Architecture for Flexibility: How to Develop a Highly Flexible Data Architecture

The internet is driving many companies to transform their business. Typically, this transformation is viewed in the context of "technical" challenges, such as how to move to an n-tier architecture, how to adopt Java and XML, and how to provide 24x7 availability. The bigger challenge, however, is architectural, and data is at the fulcrum of the issue. Why? In the early '90s, client/server technology was unproven and often unreliable. To avoid the dreaded "single point of failure", companies often deployed client/server applications as loosely coupled (stove-pipe) applications, and relied on point-to-point batch feeds to pass data between applications. The result was a data architecture characterized by highly redundant databases and spaghetti-like interfaces. Companies are now finding that the data architecture of the client/server era will not meet the needs of the internet-age. Internet protocols such as ebXML and SOAP, industry standards, and technology advances are driving entire industries to rapidly and dramatically change. To compete, companies must evolve from a stove-pipe data architecture to an architecture designed around flexibility. Fannie Mae has embarked on such an effort.

This presentation will focus on the process to develop a highly flexible data architecture. It will describe our general approach, the tools, and techniques we used. We will also discuss the design principles for flexibility and address the trade-offs we made when balancing flexibility with operational considerations, such as performance, scalability, and maintenance. Finally, we will discuss lessons learned during this effort.

John Duncan is a Director responsible for Data Architecture in Fannie Mae's Enterprise Systems and Operations organization. He has been at Fannie Mae since 1995, where he has held positions as a database administrator, technical architect, and Director of Information Architecture. He has over eighteen years experience developing database systems in IDMS, Oracle, and Sybase. He specializes in design and architecture. Prior to Fannie Mae, Mr. Duncan was a systems consultant and database administrator where he developed and deployed database systems for various Government Agencies.

How can you streamline your reporting process and reducing redundant development efforts? How can you leverage the knowledge and investment your company has spent on these reporting point solutions? Can your company move from “reports” to data warehousing and decision support?

This presentation utilizes a real-life example of a startup telecom company’s quest for information from its huge transactional databases. When your company needs everything, NOW, where do you start? The process defined in this presentation will move you from analysis paralysis into systematic, data warehouse projects through the development of a conceptual enterprise information model. Attendees will learn about a methodology that produces a framework for timely, accurate operational reporting and Decision Support activities. This process can produce efficient data stores that serve short-term reporting needs and scalable, long-term enterprise information.

Casey Welch began her data architecture and development career in 1981. She has designed, written and implemented applications for nearly every type of business. Casey was Enterprise Data Architect for a startup telecom company where she implemented the company’s decision support applications and their related processes. These efforts now provide key information that helps direct this fast-growing company in a rapidly changing industry.Currently Casey is a lead Data Architect at Hallmark Cards, Inc. where she designs solutions for Hallmark’s budding e-commerce business and its information delivery environments. Casey’s clear and creative presentation style paired with her extensive business, project management, object-oriented design and data warehousing knowledge delivers an information-packed presentation that is useful and easy to understand.

J.P. Morgan started using a metadata repository product back in 1996. Right from the beginning, we extended the underlying meta-model to capture the semantics and data structures of message formats. The emphasis was to provide the business-user view of message content to our users worldwide via the web. Message formats are stored once using business terminology and utilities are used to generate technical structures for multiple platforms. In the fall of 2001, we built an XML-enabled interface to the object-oriented modeling tool, Rational Rose. Developers can now design formats using UML and browse and reuse repository formats directly from the Rose environment.

Mary Hawley is a senior Information Architect at J.P. Morgan, a global investment bank. She has been involved in metadata projects there since 1997. When a repository product was first introduced at the firm to support Information Architecture work, Mary was the lead analyst to extend the tool to capture message metadata. Later, she became the manager of the corporate Metadata Repository. In addition to metadata management, Mary has a technical background as a programmer, database administrator, data administrator and logical modeler.

Active Meta Data as an Enabler of the ODS within an Enterprise Data Architecture

This presentation explains how the data extract, integration, transformation, and update processes for an ODS can be architected to achieve integration of data design, active meta data, passive meta data, and meta data driven processing. When developing an in-house solution for ODS processing, meta data driven processing offers significant advantages over traditional custom application code solutions. The presentation draws on the experience gained from the successful implementation of a meta data driven ODS architecture at New York State Electric & Gas Corp. (NYSEG).

Len Dawson has been a senior project analyst in the applications area for over 20 years. He has been involved in application development, production support, software package integration, development of a data warehouse, and client training. As a member of the data administration group he was the principle process architect for the successful deployment of a meta data driven ODS.

Jerry McConnell is a senior analyst in the Information Management (Data Administration) group at New York State Electric & Gas Corp. (NYSEG) and a member of the board of directors of the Rochester/Upstate New York chapter of DAMA. With over 22 years of experience in Information Systems, the last 7 years in the Information Management group, Jerry has worn many hats as programmer, analyst, project lead, client trainer, and data administration analyst. For the past 3 years he has been involved in all aspects of the architecture, design, and successful implementation of an ODS at NYSEG.