Business Mapping and ELM Overview
Effective data management starts with mapping your organization through interactive, facilitated, business driven, design sessions. The input of business representatives, and not simply IT people, is key in the ELM workshops and business interviews. Business representatives are the people in the organization who can conceptualize what is important in the organization, i.e. their core business concepts (CBCs). They are able to describe how these concepts relate to each other via their natural business relationships (NBRs). The result of this process is a fully documented ELM model, with associated artifacts, that serves as the foundation of the enterprise data warehouse.​
_edited.jpg)
An ELM facilitator will lead a series of workshops with business representatives to establish the CBCs and their NBRs. These concepts will be listed and described using the ELM artifacts.
​
The purpose of this guided process is to have a documented overview of the CBCs and their NBRs, which yields a recognizable logical data model for the business. This logical data model, or ELM, is a direct blueprint for any of the ensemble modeling patterns (such as Data Vault, Anchor, and Focal Point). It is also an enterprise-wide business model that agilely maps the business concepts within a given organization into an adaptable model.
The goal of the ELM is to map the organization. Using the ELM artifacts created in the workshop, the full logical model can be developed. This is a backbone model mapping the CBCs and how they are related. All of this is done according to the business representatives and not source systems or reference models. We use this approach because there are often multiple source systems utilized by an organization and the reference models may be too generic or abstract. To assure the meaning of the concepts is commonly understood, some attribution (context to the core business concepts) is captured, as well.
​
The ELM is input for a technical implementation of any of the ensemble modeling patterns mentioned above, or found elsewhere. Together with the artifacts, it is useful as a form of communication across an entire organization.
-
For business representatives – It is a representation of their own business, using their own terminology and their business concepts.
-
For data analysts or data stewards – It is used to understand and recognize the terminology, relations, and definitions/descriptions in their organization.
-
For source systems – The model is the map containing all the needed information (CBCs, NBRs, and attribution) available in the source systems.
-
For developers – It is an almost one to one translation towards their desired ensemble modeling pattern.
ELM Workshopping
A facilitator will lead the workshops and invite the business representatives to join the process. Each workshop starts by drafting a CBC list utilizing the knowledge of the business representatives. The next step is to categorize these concepts into event, person, place, thing, or other. Within each category, concepts are analyzed to determine if they are unique, or if they may be a synonym. To help guide the process, the group should make use of the CBC form and CBC canvas. With the connections, the NBR matrix and NBR form help to guide the team in uncovering the unique, specific, natural business relationships. These are then documented and refined for grain and cardinality.
​
On average, it will take three to six workshops with a duration of two to three hours each to model and document a given subject area. This forms the basis of an ensemble logical model. The facilitator and scribe of these workshops will need an additional eight to twelve hours per workshop to process the input into the artifacts and a base ELM model. The finalization of the CBC forms and NBR forms will be done after the workshop and it is the responsibility of the organization to implement the final model. At this point, it is likely the data steward will play the main role in this process since it involves the formalization of the business definitions and attributes plus sample data on the NBR forms.