ACWP Methodology Attribute Management: Difference between revisions

From IHE Wiki
Jump to navigation Jump to search
New page: IHE White Paper on Access Control == Attribute Management == As attributes have been assumed to be the common currency of the considered authorization ...
 
 
Line 12: Line 12:
Based on these atomic properties of the authorization problem, the needed abstractions can be built in a straight-forward manner.
Based on these atomic properties of the authorization problem, the needed abstractions can be built in a straight-forward manner.


###Classification of Attribute Stubs###
===Classification of Attribute Stubs===
###Domain Assignment###
===Specification of Attribute Value Sources===
###Policy Assignment###
===Domain Assignment===
###Specification of Attribute Value Sources###





Latest revision as of 08:42, 20 January 2009

IHE White Paper on Access Control

Attribute Management

As attributes have been assumed to be the common currency of the considered authorization problems, the following abstractions suffice in constructing the authorization solution:

  • Attribute: a distinct characteristic of an entity such as a subject or an object comprising an attribute value (e.g. 1979) and metadata such as attribute identifier (e.g. urn:example:year-of-birth), data type (e.g. integer) and category (e.g. subject)
  • Attribute value template: instructions on the construction of an attribute value (e.g. search for LDAP attribute yearOfBirth in user account identified through UUID)
  • Attribute template: attribute value template (instead attribute value) plus metadata. The attribute value template may be absent in an attribute template. The term attribute stub serves as a short-hand for attribute templates with absent attribute value templates.

The authorization system is assumed to be capable of processing arbitrary attributes and hence allowing the configuration of attribute value templates and attribute templates for arbitrary attributes. To resolve a concrete authorization problem, the first step is to identify its denominators (which may vary from one authorization problem to another) in form of a set of attribute stubs comprising e.g. the yearOfBirth attribute stub comprising an identifier (e.g. urn:example:year-of-birth), data-type (e.g. integer) and category (e.g. subject). According to empirics, the number of so-called attribute stubs happens to be rather small (say 10-25) for most practically relevant authorization problems.

Based on these atomic properties of the authorization problem, the needed abstractions can be built in a straight-forward manner.

Classification of Attribute Stubs

Specification of Attribute Value Sources

Domain Assignment


Discussion

place issues to be discussed among the editorial team here...

Change Requests

place your change requests here...