The Problem: Semantic Inconsistency
When individuals speak, it is common for Synonyms, Homonyms and Homographs to be part of the language that is used when expressing what has taken place, what should take place, or what will take place when discussing the business at hand. It is for this reason that different terms are used as Individuals and Entities (i.e., Company Employees, Vendors, Regulatory agencies (e.g., National Institute Of Standards And Technology) etc.) provide Business Requirements, Business Policies, Applications and Data Repositories to support the needs of businesses. When different Business Terms are unknowingly used to represent the same item of information, a phenomenon occurs, known as Semantic Inconsistency.
Further complicating issues is the fact that Organizational Units tend to be apportioned, and individuals are typically only responsible for taking care of matters within their own respective Organizational Units. When individuals in different Organizational Units unknowingly or knowingly perform redundant business and IT processes on the same information, they perpetuate the notion of ‘Silos’. Commonly, this notion is referred to as the ‘Silo-Effect’. In short, the Silo-Effect is the antithesis to ‘Interoperability’.
Data Models only provide a partial solution to this problem. Data modeling styles vary based on usage paradigms (e.g., Decision Support Models, Data Warehouse Models, Data Vaults). To the casual observer, the style of modeling information may seem like the ‘flavor of the day’ as Data Modelers and Data Architects come and go in a given organization. Moreover, it is not uncommon for Application Developers to be responsible for developing a Data Model when it is not their primary area of expertise. In this case, the Data Models may be incomplete (e.g., ‘completed’ without business relationships, definitions).
Narratives containing Application Documentation and Business Requirements do not solve this problem because each author is likely to have his/her own vernacular. Also, narratives often contain Business Terms, but do not effectively illustrate how the Business Terms are related to one another.
There is a direct cost associated with Semantic Inconsistency – companies that are embroiled within the ‘Silo-Effect’ are often hemorrhaging money and don’t even know it.
The Solution: GITS Taxonomy/Ontology Management
GITS has successfully utilized Data Governance to address this problem. We provide training on this subject and resources which can help you to:
- Eliminate the Silo-Effect
- Promote Synergy and Interoperability
- Reduce the number of meetings and amount of time in meetings that is typically required to understand your data needs
- Provide an efficient means of communication which can be used as a ‘springboard’ for developing Semantic Data Models and consistent Application Documentation/Business Requirements