Schema Specification of Defeasible RuleML Version 1.0

Nick Bassiliades, Guido Governatori, Benjamin Grosof, Grigoris Antoniou,
Harold Boley, Adrian Paschke, Tara Athan, Efstratios Kontopoulos, Monica Palmirani

Version history:

2007-11-09 - Version 0.9
2008-08-24 - Version 0.91

2012-07-22 -

Version 1.0

Latest version:

This is the specification of Defeasible RuleML 1.0.



Based on previous work including Courteous Logic Programs, Defeasible RuleML was first exemplified in the 'Rationale' RuleML paper and later revised for RuleML 0.87, 0.9, 0.91, and 1.0.

Defeasible RuleML 1.0 uses (control-)relationship Atoms applying a binary control relation

<Rel iri=""/>
to rule labels, where is a reserved IRI representing overriding between the labeled rules. Such overriding superiority is 'hard' in the sense of, e.g., committed choice languages, where failure of subsequent computation will not lead to a revision of a choice between rules (e.g., there will be no backtracking to an earlier Overrides choice). In an extension of Defeasible RuleML 1.0, a 'soft' variant of superiority allowing such revision could be introduced by replacing the IRI, for (irreversibly) overriding superiority, with, for (reversibly) dominating superiority. Similarly, other extensions can introduce (RuleML-internal or -external) reserved IRIs for expressing further superiority-variant semantics such as time- and/or location-sensitive overriding. The superiority relationships can be defined both by facts and rules, the latter permitting conditional superiority. In Defeasible RuleML documents in which only unconditional overriding is needed, an Overrides element
<Overrides> ... ... </Overrides>
can be used as a special atom that shortcuts the regular Atom element
<Atom><Rel iri=""/> ... ... </Atom>

applying as its relation.

Introductions to (earlier versions of) Defeasible RuleML were given, e.g., in the context of business contracts and DR-DEVICE.


Various sample Defeasible RuleML documents have been prepared and maintained; some exemplify features of Defeasible RuleML and are useful didactically while others are mostly for testing puposes.

A use case of Defeasible RuleML has been developed at Carlo-rbase-1.0.ruleml (cf. wiki history). The 'Carlo use case' version accessible to validators is being kept at carlo-rbase.ruleml.


The schema specification of Version 1.0 has been created at defeasible.xsd and dr-device.xsd.


This specification is being developed in collaboration with OASIS, in particular the LegalRuleML TC.

Site Contact: Nick Bassiliades. Page Version: 2012-07-31

"Practice what you preach": XML source of this homepage at defeasible.xml;
transformed to HTML via the adaptation of Michael Sintek's SliML XSLT stylesheet at homepage.xsl (View | Page Source)