Data Analysis & Modeling for Building the Data Warehouse

By William Shackelford

Dallas, September 10-12, 1997

What's Happening

Data Warehouse projects carry with them a problem - which is better: to try to build a perfect Data Warehouse the first time or to build a Data Warehouse that will evolve as the business evolves? Time and time again, industry experts have determined that an iterative, heuristic approach is the key to a Data Warehouse that is truly useful to the business. How can you perform adequate analysis and model the results while planning to be in a state of constant change? The Data Modeling techniques taught in this workshop will address this very issue.

About This Seminar

The activities of creating and supporting a Data Warehouse must be based on careful analysis of the data if it is to be successful at meeting the business needs. A Data Warehouse carefully constructed can be a corporate asset that gives the business a competitive advantage.

Developers can use relational data modeling techniques enhanced by newer heuristic data warehouse design techniques to guarantee that this happens. This workshop will answer the following questions:

· How can our analysts build Data Warehouses that are tightly aligned to the needs of the business?
· How can we ensure data consistency and reduce data redundancy across our Data Warehouse projects?
· How can we maximize our use of relational technology when implementing Data Warehouses?
· How can we plan for evolution of a Data Warehouse?
· How can we balance summary data with performance requirements?

What Makes This Seminar Unique

Participants in this seminar will apply the learned data modeling techniques to a Data Warehouse case study. They will experience the difficult choices that must be made in a Data Warehouse project including balancing conflicting client needs, choosing a phased out, iterative implementation, and creating a design that will be able to be under constant change. This workshop will be 20% lecture and 80% doing.

Who Should Attend & What You Will Learn

IS developers responsible for the analysis, design, development, implementation and administration of Data Warehouses will learn…

• How to convert a relationship Data Model to a Star Schema for maximized querying
• How to make significant choices between performance and flexibility in Data Warehouse design
• How to tap on existing documents to short-cut Data Warehouse design

IS Development Managers responsible for Data Warehouses will learn…

• The time and resources needed to do an adequate Data Warehouse design
• How to make business choices between performance, usability and maintenance concerns for Data Warehouse
• How to manage the challenges of loading Data Warehouses with clean, accurate information

IS Data Warehouse customers will learn…

• How to create Business Rules to ensure that a Data Warehouse is solving an important business problem
• How to validate a Data Warehouse design
• Realistic expectations about performance, querying and maintenance of Data Warehouse

Seminar Outline

1. What is a Data Warehouse?
   1. Defining a Data Warehouse
   2. How is Data Warehouse different than executive information systems, decision support systems and traditional development?
   3. The development life cycle of a Data Warehouse
2. Discovering Data Warehouse Events
   1. What is a Data Warehouse event?
   2. Discovering events
   3. Analyze the scope
   4. Brainstorm external agents’ events
   5. Brainstorm temporal events
3. The Data Model
   1. Data analysis and entity relationship models
   2. Data analyst
   3. The components of a data model (Bachman)
   4. The data entities
   5. The relationships
   6. Relationship degree (cardinality)
   7. Top down data modeling
   8. The Chocolate Factory: Relationship Matrix
   9. The Chocolate Factory: The Entity Relationship Model
   10. Is it a data entity?
4. The Progression of Detail
   1. The importance of the repository
   2. The Basic Entity Relationship Model
   3. The Keyed Entity Relationship Model
   4. Replace the many to many relationship
   5. Adding an associative data entity: example
   6. Determining foreign keys
   7. Compound keys vs. Indexing
   8. The Attributed Entity Relationship Model
   9. Normalization
      1. Why normalize?
      2. Normalization: first normal form (1NF)
      3. Normalization: second normal form (2NF)
      4. Normalization: third normal form (3NF)
      5. Normalization: adjust the attributed model
      6. Normalization: example
   10. Final considerations
   11. Refinement
   12. Evaluate the one to one relationship
   13. Evaluate the one to one: example
   14. Remove redundant or incorrect relationships
   15. Removing redundant relationships: example
   16. Adding existence criteria
   17. The attributed model with existence criteria
5. Data Warehouse Trade-offs
   1. Different views of data
   2. Logical and physical synchronization
   3. Types of data: metadata
   4. Adding physical information
   5. Time series data
   6. Performance considerations
   7. Multidimensional modeling
   8. The star schema
   9. The fact constellation schema
   10. The snowflake schema
   11. Tips for aggregation
   12. Drilling down
   13. Identifying the source
   14. Data replication
6. Other Considerations
   1. The appropriate use of CASE
   2. CASE should provide
   3. Operating in crunch mode
   4. Other Models

About Your Instructor

William Shackelford is a Senior Learning Facilitator for Russell Martin & Associates, a firm specializing in timely technology challenges such as Data Warehouse, Year 2000 and Project Management. He has also been President of Shackelford and Associates since 1982. He completed his undergraduate and graduate studies at Indiana University, Universitat Hamburg, Illinois Institute of Technology and the University of Illinois in music, accounting, business administration and computer science. His extensive background in business systems development involving systems analysis, design, testing and support experience has brought success to his customers including Amoco, Kemper, R.R. Donnelley, Avon and Dean Witter.

Meeting Site and Hotel Information

Dallas, September 10-12, 1997
Doubletree Hotel at Campbell Centre
(214)691-8700

Register Now!

Data Analysis & Modeling for Building the Data Warehouse

$1195

Attend this seminar and Managing Data Warehouse Projects: The Key Issues and SAVE $395

ON-SITE SERVICES

Bring This Seminar to Your Facility!

How Does On-Site Training Benefit You?

1. Save on your travel budget
2. Focus on your specific needs
3. Customization
4. Train your team together
5. Convenience
6. Confidentiality
7. On-going consulting services

Call (508) 470-3870 today for more information and ask for the training project manager to keep your company ahead of your competitors.

DCI ON-SITE SERVICES

Education Developed EXCLUSIVELY for Your Organization

Stretching your training dollars is more important now than ever. Training for changing technology can be costly and time consuming . . . but with DCI it won't be.

DCI, the world leader in IT education, can bring our experts to your facility. DCI will meet your budget and reach your training goals.

Call (508) 470-3870 or Email mmullen@dciexpo.com today for your free quote.