Architector Semantic Business Glossary

A semantic glossary of business terms is much more than a simple dictionary of data. It contains definitions for terms used in the business, certainly, but it also:

  • Classifies terms according to the type of concept represented (i.e. noun, verb, categorization, instance, data type, data property, business rule, nominalization).
  • Defines meaningful relationships between terms, i.e. types, aggregates, categories, verb associations, property of, instance of, synonym of, rule for, and roles.
  • Maintains useful metadata for terms, e.g. Source of Definition, Definition Status, etc.

Architector’s semantic business glossary has all the features listed above, and many other unique and insightful features.

 

Component Architecture - Glossary

Architector Component - Glossary

Semantic Parser

Architector automatically parses every definition and most semantic relationships between terms are established directly from the term definitions. Architector then draws the corresponding concept model directly from the parsed definitions. This approach has major benefits:

  • The concept models are always an accurate representation of term definitions.
  • Concept models do not need to be manually drawn, and do not need to be re-drawn when a definition changes.
  • The automatic visualization of definitions provides a powerful feedback to the definition team, and helps to establish a common understanding.

Concept Visualisation

Architector can automatically visualize the semantic relationships between concepts. For example, here is a definition of a verb in the glossary:

Component Glossary Def

The parser has recognised three noun-type terms, underlined in green. And also one verb-type association, in blue italics, between two of the terms. In the Structure diagram (concept model) this is displayed as:

Component Glossary Concept

Definition experts and modellers may well argue that this representation is not correct. It may be better to define a noun: Assessment of Risk for a Customer, and to record this in the Customer Risk Assessment. If we agree then all that needs to be done is to change the appropriate definitions, and the model will be automatically updated.

Definition Scoring

You will notice a score given to the definition in the above example. Rules can be defined within Architector, by which the quality of definitions are measured. The system comes with many automated rules, implemented using natural language processing (NLP), that can be turned on or off. And additional rules can be added. Architector automatically assesses new and changed definitions. This is very useful feedback to a definition team, who can now objectively define and manage the quality of their data definitions.

Semantics of Business Vocabulary Rules (SBVR)

SBVR is a publically available standard for defining concepts, associations, and business rules. Architector’s parsing and concept management system is largely consistent with SBVR, and uses many of the SBVR conventions. This provides a common language to define concepts and data. However you are not forced to use SBVR-style definitions should you prefer a different approach.

Vertical Lineage

Architector maintains the links between business terms and actual data usage in processes and systems, and these links are accessible via the business glossary. In the example below the data element Customer Risk Level is stored in three datastores: Customer Risk File, Aggregate Sales Report, and Customer Risk Assessment. We can also see from this view the physical data name, the data type, and how many processes the data is an input or output to.  And we can see if the representation in a datastore is the ‘golden source’ for the term, and if there are any known issues associate with the term in the datastore (e.g. data quality issues).

Component Glossary Tabs

Other powerful views are also available from this screen by using the different tabs. Please contact us to request a demo or for more details.

Upgrade your glossary, and unleash the potential of Architector!

Case studies:

Due to the nature of work we undertake, and the types of clients we typically help, we do not post details of case studies online. Please contact us for more details.

×