A business glossary is a major component within a data governance (or data intelligence) initiative. Glossary entries consist of a name and a functional definition. The purpose of a functional definition is to explain the meaning of a piece of data in such a way that anyone within the organization could understand the meaning of that piece of data. A glossary entry is different from a data dictionary item. A glossary entry has more narrative, and a single glossary entry may cover several dictionary items. For example, the glossary entry would be Employee Legal Address and it would cover data dictionary items such as Address Line 1, City, State, Zip Code, etc.
Organizations need hundreds of glossary entries to comprehensively explain their data, and the functional definition needs to be written fairly quickly. How do you train staff to write good definitions? What is a good definition? Typically definitions are written by multiple people, none of whom are lexicographers by training. At IData we teach a method for writing functional definitions that is easy to learn and to use. Adopting this method prevents circular or vacuous definitions and it provides a framework that ensures consistency across multiple writers.
This method contains 6 components. Components should be tackled in the order listed. Components with the asterisk are required. The components are:
- Uniqueness within a category OR Description*
- Re-use definitions within definitions*
- Add business rules
- Add valid values or reference data lists
- Add quality rules
- Edit the definition for the perspective of the smart outsider*
Let's apply each component to the definition of "Major".
Name: Major
Functional Definition: A field of study within an academic discipline that a student has chosen to pursue. A student must pursue a primary major and may optionally pursue additional majors. To earn a degree, a student must meet all major requirements. The list of majors is maintained by the Registrar's office and approved by the Provost. A student can declare up to 4 majors.
The first sentence identifies the category "field of study", and then differentiates this field of study from all others fields of study by the qualifying phrase "...that a student has chosen to pursue". This sentence pertains to the first component, which calls for identifying a category and then identifying the uniqueness of this item from all other items in the category. About 70% of definitions can use this categorization and differentiation method. For the remaining 30% that have no meaningful category, describe the item, using common sense to guide the necessary level of detail. A subsequent blog post will provide more examples for this component, since it is the most difficult component to grasp.
The second component is illustrated by the underlined words or phrases. Each underlined word or phrase should be defined in the glossary as well. Re-using definitions in this manner discourages the use of synonyms or similar-meaning words, which introduces ambiguity. The Data Cookbook provides URL links to these definitions so that the user can quickly navigate to learn the meaning of terms that they may not understand. This practice also prevents the inclusion of further definitions within a definition.
The second and third sentences provide business rules that explain how a Major is selected and earned. These sentences illustrates the third component in the method.
The fourth sentence pertains to the valid list of majors and corresponds to the fourth component in the method. It indicates the owner of the list (such as a Registrar's Office), and identifies who has the authority to approve the list (such as a Provost).
The final sentence infers a quality rule, which is the fifth component. By indicating that a student can select multiple majors, a data consumer is given notice that the number of majors vs. the number of students may differ.
Finally, the last and sixth component asks the writer to edit the entire definition from the perspective of a smart person who is not familiar with enrollment data. The definition is free of technical jargon and is grammatically correct.
This method is adopted from Malcolm Chisholm's book Definitions in Information Management : A Guide to the Fundamental Semantic Metadata. We have adopted the components most critical and relevant to higher education. Use IData's method to train staff within a few hours. These 6 components are easily memorized by staff who do not write functional definitions full time.
If you need help in implementing data governance, remember that IData provides data governance services. Other resources regarding data dictionaires and definitions can be found in this blog post. A data governance solution like the Data Cookbook can help in successful implementation of data governance at an organization and improving data quality. Feel free to
.
(image credit StockSnap_FYLN4CE6C7_PenPaperKeyboard_WritingGoodDefs_BP #1042)