Glossary

attribute - (in a logical data model) a property or characteristic of a logical entity, which is qualified by the value assigned to it - e.g. for attribute = 'Eye Colour', value = 'Blue'.

business intelligence - information which represents a historical view of the output from business activities, and how it changes over time.

cardinality - (in a logical data model) a property of a relationship between two entities which specifies how many of one entity can be related to how many of the other entity - e.g. 'one-to-many', 'many-to-many', 'one-to-zero-or-one', 'one-to-four' etc.

commonality - (with regard to data) an instance of two physical data elements, or combinations thereof, having the same logical definition.

constraint - (in a logical data model) a condition (or limitation) which applies to any instance of a particular entity, attribute, domain value or relationship between two entities, and which cannot necessarily be fully illustrated by the notation used to document the logical data model. Note that a constraint which cannot be handled by the modelling notation is typically described in the relevant definition.

data dependency - an instance of a business activity creating, reading, updating or deleting a data element (both logical and physical).

data profiling - the process of examining data by performing statistical analyses of data values & formats.

data warehouse - one or more databases which constitute(s) a data storage application to facilitate the provision of business intelligence.

disparity - (with regard to data) an instance of two physical data elements, or combinations thereof, having different logical definitions.

domain value - (with regard to data) a character string which belongs to a set of permissible values which may be applied to a particular attribute, especially with regard to reference & master data.

entity - (in a logical data model) a concept which is involved in business activities - e.g. 'Customer'.

ETL - Extract, Transform & Load - the three stages involved in the integration layer of a data warehousing process - i.e. extract data from a source system, transform it so that it can be integrated with data from other source systems, then load it into the data warehouse.

flat file - an unstructured source of data.

foreign key - one or more attributes within an entity whose values constitute the relationship to another entity (because they equate to the values which uniquely identify an instance of the related entity).

form - an object within a software application which allows a user to create, read, update and delete data in a particular data storage application (depending on a set of specified permissions).

integration layer - see ETL.

logical data model - a form of notation which defines the data structures involved in a specific business domain and is expressed independently of any particular data storage application.

master data - attributes (usually of low volatility) which represent business concepts which are agreed on and shared across the enterprise - e.g. 'Customer Name'.

metadata - data which gives context & meaning to other data, and is therefore present in captions on forms, column headers on tables etc.

optionality - (in a logical data model) a property of a relationship between two entities which specifies whether or not an instance of one entity must have one or more (depending on the cardinality) related instances of the other entity.

persistent - (with regard to data) does not change over time.

physical data model - a form of notation which defines the data structures involved in a particular data storage application.

primary key - one or more attributes within an entity whose values uniquely identify each instance of the entity.

RDBMS - Relational Database Management System.

reference data - attributes (of very low volatility) used to define the set of permissible values which may be applied to qualify other attributes - e.g. 'Unit Of Measurement', 'Country Code'.

relationship - (in a logical data model) an association between two logical entities represented by one or more common attributes.

structured methods - a set of techniques used to analyse and document the nature of business activities.

subject area - (in a logical data model) a collection of entities, and the relationships between them, relevant to a particular business topic.

volatility - (with regard to data) the rate of change of a particular set of data.