The use of agglomerative categorisation to build a conceptual view of statistical data

Abstract

The creation and use of the Divisional model provided a means to record the data components of the Division's systems.This made it easier for the data analysts and designers:

• to develop new application data models rapidly,

• to learn and remember the overall structure of the Division's conceptual data,

• to enforce centralised modelling constraints,

• to train new developers

• to participate effectively in data modelling activities,

• to identify reusable data structures, and

• to identify and eliminate duplicated data structures and related processing.

Although it was developed primarily to assist in the migration process it was also used effectively in the development of new systems and has continued to be used in this way to the present day.

If this principle of agglomeration is portable to other domains then the experience of Statistics Division suggests that it could be used to speed up the development of corporate or enterprise models, improve their effectiveness, utilise the skills of less experienced modellers and significantly increase productivity.