DAMA + Wilshire Meta-Data Conference - Data Strategy Track
Diagrams play a critical role in data management practice: they are used to document information requirements and architectures and communicate these to stakeholders throughout the organization. Despite this, data management professionals typically receive little or no training in how to produce “good” diagrams. As a result, they are forced to rely on their intuition and experience (which is often wrong), and make layout decisions that distorts information or conveys unintended messages. The unfortunate but inevitable consequence of this is that most diagrams used in data management practice communicate very poorly. Although they are intended as a way of communicating with end users and senior management, they more often act as a barrier rather than an aid to communication. This workshop describes a set of principles for producing “good” diagrams, which are defined as diagrams that communicate effectively. These are based on evidence drawn from a wide range of fields, including visual perception, cognitive psychology, graphic design and diagrammatic reasoning. The principles apply to all types of diagrams, from formal diagrams used in application development and enterprise architecture to informal diagrams used in presentations and reports.

What you will learn from this workshop:
  • What is meant by a "good"diagram and how this can be measured
  • Common errors in diagramming practice and how to avoid them
  • The language of graphics: learn the full "vocabulary" of techniques for graphically encoding information and how to use them effectively (laws of graphical composition)
  • Graphical information processing: learn how the human mind processes graphical information and how to use this knowledge to develop diagrams that optimize understanding (laws of perception and cognition)
  • Principles for producing effective diagrams: practical guidelines for producing diagrams that communicate effectively with all stakeholders, especially those from non-technical backgrounds (i.e. end users, customers and senior management)
Ever ski? Each trail is graded a green circle for easy, a blue square for moderate, or a black diamond for difficult. The modeling challenges you face in the office can be put into similar categories.

This workshop includes a carefully selected collection of easy, moderate, and difficult scenarios. After mastering easy challenges, you'll advance to more moderate and difficult challenges. This is not just a lecture. You'll all get hands-on experience. The more times you fall and pick yourselves up again, the more trees you hit, the more you challenge yourself, the more you'll grow your experiences and knowledge base.

Just like skiing, expect a fun yet intense day. By the completion of the workshop, you might be sore and sweaty, but you'll obtain a higher level of experience and have a few more techniques to apply when you get back to the office.

This workshop includes three modules. Module 1 contains the easy Green Trails, Module 2 the moderate Blue Trails, and Module 3 the difficult Diamond Trails.
  • Green trails will strengthen our skills in areas such as business rules and assertions, normalization, nullability and definitions.
  • Blue trails will strengthen our skills in areas such as abstraction, data politics, reverse engineering, surrogate keys, and summary tables.
  • Diamond trails will strengthen our skills in areas such as dealing with unrealistic timeframes, history, integration, BCNF/4NF/5NF, and dimensional modeling.
“If you obey all the rules, you miss all the fun.” - Katherine Hepburn

This SIG will focus on a discussion of CA’s AllFusion ERwin – its functionality as well as tips and suggestions to enhance the tool’s productivity. Facilitated by the President/CEO of Axis Software Designs, a Model Management Services and Education company specializing in AllFusion Product Suite training and model management infrastructure consulting, this SIG should elicit personal experiences of tips and shortcuts, as well as ‘Don’t Dos!’ from the audience. Come to learn, and come to share.
This presentation describes a framework for creating domain models using four colors to stereotype the entities (or classes for those that prefer the UML modeling notation) into five categories. This speeds up the analysis modeling process, produces models that are easy to review and critique, and provides a consistent way of thinking about items of interest to the business. The use of color adds to the information content of the model and provides an excellent mechanism for visual discrimination of the concepts.

The presentation begins with a brief description of the five stereotype entities, followed by an example, tips for identifying those entities, typical attributes in each entity, and concludes with a stereotyped way the stereotype entities fit together.

To make the concepts concrete, one of the included example models is of this year’s DAMA International Symposium & Wilshire Meta-Data Conference!

After the session, attendees will be able to:
  • recognize when to apply the five stereotypes
  • apply the four colors in your models
  • think about the world differently.
ORM is a better, richer way to do high-level, conceptual data modeling.  It is distinctly different from the more traditional,
record-based schemes such as ER, Extended ER (EER), IDEF1X, IE, or UML.  It was originally developed by Shir Nijssen (called NIAM) and further enhanced by Terry Halpin.  ORM is now embodied in Microsoft's Visual Studio.net Enterprise Architect edition, and called VisioEA.  However, it is practically impossible to use that tool without some basic understanding of ORM.

This presentation covers:
  • The essence of ORM data modeling
  • Relationship of ORM to Object-Oriented Design and Development
  • Comparison of ORM to conventional record-based modeling schemes such as ER
  • Transitioning understanding from ER/Relational to ORM
  • The basic constructs of ORM - the elementary fact sentence with one predicate (a relationship) and one or more objects (the entities)
  • Representing a rich set of semantic constraints in ORM
  • Dependency, uniqueness (multiplicity/exclusivity), value sets, role populations, frequency, ring...
  • Graphical and verbal representations of an ORM model; abstractions in ORM
  • Mapping an ORM model to a Relational data model
  • Automatic generation of a fully normalized model from an ORM model
  • The place of ORM in a taxonomy of data modeling schemes
This talk is a must for anyone involved with logical data modeling. It behooves every data professional to consider the ORM data modeling scheme and to critically examine the claims made for it. Find out for yourself if it really is a better way to do data modeling.
In information systems modeling, the business domain being modeled often exhibits subtyping aspects that can prove challenging to implement in either relational databases or object-oriented code. In practice, some of these aspects are often handled incorrectly or inefficiently. This presentation identifies a number of subtyping issues that require special attention, and shows how to model them conceptually, and then map them to both relational and class structures for implementation. To cater for different preferences in modeling approaches, the examples will be portrayed using Object-Role Modeling, Entity Relationship, and UML notations, as well as textual languages for rule verbalization.

The main topics discussed are:
  • Quick Review of Basic Subtyping Principles
  • Subtyping Constraints beyond Disjoint and Complete Restrictions
  • Asserted, Fully-derived, and Semi-derived Subtypes
  • Role Subtypes vs Sortals
  • Mapping Subtyping to Relational Databases
  • Mapping Subtyping to Class Structures (OO code)
More industries have published standard data models than ever before. These models, often provided for free or only minimal cost, can add a great deal of value to your modeling efforts. In this presentation, Karen discusses some of the benefits and gotchas of working with acquired models - industry standard models. This session includes topics such as:
  • Survey of industry standard data models
  • Costs, benefits, and risks of working with industry standard data models
  • Using the right process
  • Myths in working with pattern models
  • 10 Tips for successfully working with third party models
    • Lessons Learned

OMG’s Common Warehouse Metamodel (CWM) is mature and stable, with widespread and still-increasing adoption by vendors and customers for metadata interchange: most widely in the area of relational database information.

OMG is now in the process of adopting a replacement standard called the Information Management Metamodel (IMM) that broadens the applicability of the standard and its integration with a number of other OMG standards to address many other areas including UML, Ontology Modeling, XML Schemas and Service Oriented Architectures; and the automated transformation and management of development through OMG’s Model Driven Architecture approach.

Another key aim of IMM is to increase take-up and tool support in the data management community. At the time of the DAMA conference the submission will be at initial revision stage and one aim of the presentation is to solicit feedback and involvement.

This presentation will cover the following aspects of IMM:
  • Background
  • Requirements
  • Usage scenarios
  • Overview of initial submission
Bigger picture of OMG standards:
  • Technology support including Eclipse
  • Roadmap
  • How to contribute
If you are engaged in data modeling efforts, whether large or small, your ongoing success is dependent upon a structured, well planned model management environment that is focused on reusability and non-redundancy.

In a multi-user, multi-project environment, your modeling strategy should drive the development of a structured approach, implemented by a supporting infrastructure. A repeatable, efficient modeling life cycle is critical to effectively managing the environment, as well as receiving the most possible return on investment for your modeling efforts.

Most organizations have a baseline set of standards for building logical and/or physical models, but lack the critical ‘reusability and redundancy analysis’, as well as a host of other necessary procedures to implement a strategic approach to modeling. No matter what data modeling tools are being used, a formal infrastructure is the only means by which consistent modeling projects can produce model structures with integrity and reuse potential.

This session will discuss the importance of establishing a model management strategy and will highlight the key factors involved in prioritizing and developing a successful infrastructure.

Content Highlights The Big Picture
  • Let’s Hear It For An Infrastructure
  • Strategy First
  • Without City Planning, It’s Only Suburbs
  • Reusability (Saving Time & Money)
  • Non-redundancy (Haven’t We Done This Already?)
  • Metadata integrity (Who Can You Trust?)
  • Then Implementation
  • What Do We Want To Achieve? The Plan -Purpose - Strategy
  • Plan Details - Model Development Life Cycles - Standards
  • The Critical List - Procedures
  • The Critical List - Prioritization is Vital
  • It Doesn't Happen All At Once
  • And Finally… Communication Is Everything
We all use mental maps of how our world is organized to accomplish everyday tasks from opening doors, stopping at traffic lights, reading a newspaper to adjusting the water temperature in the shower. They define our sense of how things should be. If we lay out data models with an understanding of the these maps, we make it easier for our audience to see what we want to communicate.

Mental maps are simple to understand and identify once we realize they are there. But they are also very powerful in their effect. We will explore visual maps and how they translate into simple layout guidelines. Some of the maps we will cover are:
  • time lines put the early events on the left
  • hierarchies go from general at the top to specific at the bottom
  • white space identifies clusters of related objects
  • short lines are easier to follow
  • repeating patterns identify similar concepts
The debate about whether to use a star or a snowflake approach to dimensional modeling dates back to the early days of data warehousing. Increasingly sophisticated database systems and tools and faster hardware are changing the tradeoffs between performance and flexibility.

I will present the results of real-world tests comparing the performance of identical data marts implemented with both star and snowflake schemas.

This will feed into a framework for choosing which modeling approach is best for a given requirement. In particular, we will examine how the star/snowflake decision is driven by:
  • Load and query performance
  • Data model usability and maintainability
  • Aggregate design and flexibility
  • Ability to conform dimensions
  • Tool selection
The ER/Studio Special Interest Group (SIG) provides users with the opportunity to discuss best practices, exchange tips and techniques, and explore advanced capabilities of ER/Studio and Repository. Topics to include:
  • User experiences in use of new product features
  • Best practice sharing
  • Suggestions for new features
  • and other discussions you will not want to miss! Come meet others in the data modeling community and share in this fast-paced hour geared towards learning how to maximize the use of ER/Studio!
    Wednesday, March 7th
    5:30 pm - 6:30pm

    PowerDesigner SIG - Michael Nicewarner, John Deere Credit

The PowerDesigner SIG provides an open forum to discuss experiences and techniques with the most popular modeling tool. We will talk about:
  • The various model types (ERD, UML, BPM, XML, etc.)
  • Techniques and best practices when using the tool
  • Issues, ideas or bugs to send back to Sybase
A description of this session coming soon

 

Close Window
Wishire Conferences DAMA International