Enterprise Information Architecture at Nokia Using Universal Data Models: A Case Study
Len Silverston
President
Universal Data Models, LLC
Teemu Mattelmaki
Director, Chief Information Architect
Nokia Corporation
Nokia developed and implemented an enterprise wide information architecture to move towards integrated information versus data silos. How did they do this and what were their challenges? How did they re-use common data models to leverage their efforts and where did they need to customize these “universal” models? This presentation will share experiences and techniques that Nokia used to facilitate shared information at Nokia. For example, the director and chief information architect at Nokia will show how universal data models were applied in Nokia’s Enterprise Information Model (NEIM) and how NEIM was applied in applications using business subject area models. Nokia socialized the concepts of its enterprise model being “City Plans” and business subject areas being “City Area Plans” with the database designs being the “building blueprints”. A “think big-start small” approach was a core principle of this program.
In this presentation you will learn:
- How to effectively apply “Universal Data Models” to a large organization such as Nokia
- Examples of these models with instance diagrams of data
- The various architectural dimensions that were employed at Nokia and how they relate to each other
- Various methods and techniques used in this enterprise wide effort
- What worked and what didn’t work and why
Comments (1)
Len Silverston, Universal Data Models said
at 8:33 am on May 5, 2006
The main points to take home were:
- Using UDM as a reference model pays off.
- Different kinds of model views for different audiences aids in understanding the requirements, validating the models, getting buy in, and communicating effectively.
- Flexible solutions are based on a stable architecture that more easily allows change. Flexible data models facilitate a stable architecture.
- An architecturally solid, logically consistent, flexible,
scalable data model should be based on common constructs, a common
enterprise wide data model and related design patterns
- Further development of the data model into a physical data model
should be based on a generic enterprise data model in order to provide consistent and flexible data structures
- For business people we can use different kinds of visualizations
(e.g. instance diagrams, specific data models) in order to
show that the business requirements are covered. These models are used for
capturing requirements and creating specific views in order to understand and validate information requirements. However, these specific views do not generally form the basis for a solid, stable physical database design. A flexible, generic enterprise data model works well as the basis of an architecturally solid database design since it adapts well to changing needs.
- Nokia has used this strategy to create small successes and has gained momentum in their data management program by re-using common models and patterns.
You don't have permission to comment on this page.