Goto: Documentation Home

Databases often contain lots of information, but typically not the piece of information you want. You may have a database with names and addresses, and another with demographic data based on suburbs. The problem is how to connect the two databases together. This is where universal domains help, by informing GDS that the names and address database really does have an address, the system is able to match other databases that also have the universal domain address defined. This matching is able to understand differences in addresses and still match records.

Reference Overview

• Each universal domain is allocated a unique number/key. If this is a single number (eg 26) then this is a globally defined definition, if the key includes a dot (eg ABC45G.26) then the definition is specific to a particular environment/database.
• Universal domains do not necessarily define the exact storage format. A complete address may be stored in many different ways and univeral domains are not trying to stipulate how this is stored. Applications are free to store data in whatever format they wish
• Universal domains can refer to a closed set or an unbounded set. A closed set has a finite number of elements and it is reasonable to expect a list of members to be available. For example "country names" or "color names" are both closed sets.

  Unbounded sets are sets that are virtually unlimited in potential members (eg "prime numbers"). Many of these sets can provide an incomplete reference set however, such as "first-name", which may provide a list of members as best known at the time.

• Every member of a universal domain set is allocated a unique physical key, and once you know the physical key of a record the system can scan for additional relationships and attributes based on this physical key value.
• Universal domains implicitly categorise data and any system of categorisation will not be suitable for all situations. Consider the universal domain "country names", is this only countries recognised by the UN? Should historic countries be included? What about individual countries recognising the existence of other countries? For these situations the system remains with a single universal domain that is tightly defined as to what it describes and then additional attributes for each member provide sub-categorisation (UN member, historic, etc)

Standard Domains

Common concepts are allocated global common constant identifiers, which consist of only digits

1	city	A city or town name
2	address-line	A single line of an address
3	time-absolute	A point in time such as when something occurred
4	quantity	An amount of something such as the count of beans in a tin
5	description	A label description or tag. Generally human readable
6	id	An identifying number that in itself has no meaning
7	money-total	A sum of money
8	address-house-number	A house number being the single digit value of an address where present
9	address-street-name	The street name an address belongs too
10	latitude	A geographic latitude
11	longitude	A geographic longitude
12	first-name
13	surname
14	person-name
15	company-name
16	nickname
17	date-of-birth
18	song-title
19	book-title
20	country	Country Names
21	musical-instrument
22	barcode
23	registration-number	Vehicle registration number
24	email	Email address
25	phone	Telephone number
26	Street types (English)	List of words commonly used as street types, such as road, place, way.
27	City, suburb, region	A regional descriptor, which can be a suburb, town or city, state
28	Color-names	Names of colors
29	Prime numbers	This domain is primarily used for testing purposes
30	Barcode Symbologies	Known Barcode families
31	Packaging types	Common product packaging types, such as tin, bag, box, packet
32	Zipcode	An American ZIP code. This field only refers to US zipcodes
33	State Code	An American state abbreviation, eg IL, CA, TX