Standard Data model role in building Information Framework

Information framework in nutshell represents the ecosystem where disparate systems/components can exchange information with each other using a common underlying data structure. In this write-up, I am going to highlight the role of industry-standard data models in building an information framework and how it aligns with the organization's data-driven strategy.

Photo by Kaleidico on Unsplash

Overview

So what is an industry-standard data model all about?

1. First and foremost Industry-standard models represent a standard way of structuring, defining, and implementing an information framework.
2. It provides a consistent and common vocabulary to cover enterprise knowledge at each abstraction level.
3. It focuses on detailing the data landscape of an organization and best practices to cover the unknown requirements at the time of initiating a new request or during a change request.

What all the basic elements it covers and where does it map to Organization’s data landscape?

1. Data Elements
   It represents the metadata in a holistic manner so you can focus more on implementation rather than coverage aspects. It deals with common terms which are well understood across the partner ecosystem as well along with providing a consolidated and holistic Data Definition that helps you to jumpstart building the knowledge repository.
2. Data structure
   It represents real-world information by translating real-world objects and their state representation in various ways by using Entity based model via the concept of decomposition of Hierarchies [subtype/role play/dynamic] and classification schemes.
3. Data mapping
   At each abstraction level, data models enable you to map the information back to processes (business/technical) so it complements the business process(s) implementation in an unified manner.

Benefits

What are the key benefits it brings along with it?

1. Reduced time to market and thus expedites business solution
   In many ways, a template-based model substantially reduces the time to market for a product or feature implementation. While the common terminology aids the development/integration with the partner ecosystem the holistic nature of the data model provides coverage to all sets of requirements known/unknown. It often provides a jump start to all the teams involved in rapid prototyping the existing state of solution and also helps in doing the feasibility analysis of a new feature or solution
2. Reduces cost of integration and improves overall efficiency
   When disparate teams converge on a common vocabulary, it is often observed that it improves the overall process efficiency by minimizing the cost of integration. Further, it aids the design of a unified business logic that can be implemented once and can be re-used by other teams.

Alignment to Data Strategy

How does it align with my data strategy?

As a data-driven organization, how can I seamlessly integrate it?

In all possible means to adopt a data-driven culture, as a key enabler, the data architecture should resonate with the data strategy in principle. The data model closely aligned with key components of data strategy as outlined below in brief

Define

The industry-standard models help to identify the key elements which are critical for an organization. One can easily start with adopting the key definition to start building the basic model and perform further iterations to refine the model as per organization needs. The model helps further by acting as a bridge between business and technical people and allow them to reach a common consensus despite having different viewpoints on the same contextual representation. Having a common vocabulary plays a pivotal role in achieving this common ground of understanding.

Provision

A common structure for data information exchange enables the organization to rapid service provisioning and problem handling. This is especially vital when the organization process is dependent on the partner ecosystem for enabling a distributed value chain. Moreover, with a common vocabulary, the model quickly becomes a benchmark to perform rapid prototyping and democratizing the data.

Process

Data model helps in risk reduction by ensuring completeness which is holistic coverage of all the known / unknown requirements supplemented with industry best practices. It also helps in breaking silos by acting as a bridge to aid coordination and integration of disparate data sources thus improving overall process efficiency and reduced the cost of integration

Governance

Data models ensure the implementation of data quality control by helping in building institutional memory. With the right set of governance principles, it compels data stewards to maintain a healthy data dictionary along with metadata capture to understand the nature of data like how many distinct values, null allowed, whether a key identifier or not, etc. It aids the data provenance by allowing a map at each abstract level as well. Finally, it gives the contextual meaning to information/data along with outlining all active entities/instances involved in defining that particular context and thus acts as a key enabler for building a knowledge graph.

Agility

Data models support agility by employing key design techniques such as the Hierarchical model or Sub/Supertypes or simply a type design for capturing dynamic nature to represent the state of data. At a logical level, it essentially represents a way to traverse the entities involved which serve as a basic element to design “Analytics as a code” functionality to perform ad-hoc analysis. Often people fail to neglect the importance of having the right design when they focus too much on designs based on “schema on arrival” and completely neglect the “Schema” word in it.

I have tried to summarize all the above points in a pictorial representation as below

Data strategy acts as a guiding principle for defining data architecture as the strategy links the business goals to the technology solutions. I am convinced that industry-standard data models which are key components of data architecture, serve as a key differentiator in achieving a pragmatic balance between “what” to do and “how” to do components of data strategy.