Schema Specification of RuleML 0.89

An introduction to RuleML is given in our tutorial. Also, the main page's design section discusses the upper layer of the RuleML hierarchy of rules. In that terminology, the system of RuleML XSDs presented here only covers derivation rules, not reaction rules.

This is because we think it is important to start with a subset of simple rules, test and refine our principal strategy using these, and then work 'up' to the more general categories of rules in the hierarchy. For this we choose Datalog, a language corresponding to relational databases (ground facts without complex domains or 'constructors') augmented by views (possibly recursive rules), and work a few steps upwards to further declarative rules as allowed in (equational) Horn logic.

Below is a summary of the changes in version 0.89:

For more information, see the Changes section.

The grammar of RuleML 0.89 (i.e. the content model of each individual tag) is available as a single document called Content Models for RuleML in addition to being encoded within the XML Schemas. Issues relevant to RuleML 0.89 were already discussed in the Issues List of FOL RuleML, in particular issues 1. and 9.

A presentation entitled From RuleML 0.88 to 0.89: Sublanguages Beyond Horn Logic - Validation and Translation is also available.

See also a recent publication accepted for the W3C Rules Workshop in Washington: Implementing RuleML Using Schemas, Translators, and Bidirectional Interpreters.

RuleML 0.89 consists of the following changes relative to the previous version 0.88.

Four new sublanguages providing the modular specification of FOL RuleML have been added to the family:

Another three new sublanguages are also added to provide the serialization of the Semantic Web Services Language (SWSL):

A new tag <Reify> is introduced to support reification, a kind of instantiation or quasi-quotation. It is permitted anywhere a regular term (e.g. Ind or Var) is, in all sublanguages, and allows any RuleML tag available within the current sublanguage as content, treating it as a term for performing reasoning on. For example, the following is a reification of the 'own' rule:

buy person merchant object keep person object own person object ]]>

See the SWSL report for more details on <Reify>.

Two new attributes 'minCard' and 'maxCard' are introduced to allow cardinality constraints on slots. For example, the following specifies that the cardinality of a slot is no less than 1 and no more than 2:

... ]]>

Another important addition is <Data> that may optionally specify an XML Schema built-in datatype (e.g. date, integer, string) for checking during validation using the 'xsi:type' attribute. For example, the following are all correctly enforced during validation:

test 12 2002-10-10T17:00:00Z]]>

And, as expected,

10000000 ]]>

fails, where XSV returns the following:

element content failed type check: 10000000>32767 ]]>

Note that <Data> without an xsi:type attribute allows unrestricted text, as in RuleML Lite, and also that the following namespace declaration must be included to refer to XSD's built-in datatypes (as present in datalog.ruleml):

Finally, note that other terms (e.g. Ind and Var) may only have user-defined types via the existing 'type' attribute. Ind roughly corresponds to resources in RDF, while Data corresponds with RDF literals.

Mutual exclusion is possible with the new <Mutex> tag, allowable as the first child of the new <Protect> performative (optionally wrapped by the <warden> role introduced to maintain RuleML's striped syntax). A <Mutex> states that it is a contradiction for a particular pair of literals (known as the "opposers" and denoted by "oppo") to be inferred, given another optional particular condition (known as the "given", marked up as "mgiv"). For example:

pricediscount CUST Y pricediscount CUST Z Y Z ]]>

See Benjamin Grosof's Courteous Logic Programming for more details.

Introduced to Atoms (at hornlog), Cterms and Plex's are an optional positional rest role <repo> and a slotted rest role <resl>, both of which may only contain a Var or a Plex. For example:

person who person ageage X ]]>

With the new <Skolem> term (available in all sublanguages), RDF blank nodes can be captured as Skolem individual constants. For example, POSL's _PeterMillerPC becomes

PeterMillerPC ]]>

and Skolem generator '_' becomes

]]>

Object identifiers (<oid>) are now permitted in all atoms, connectives and quantifiers, now available for the first time in all instances of And (not just directly under Assert), Or, Naf and Neg.

The 'innerclose' attribute is renamed to 'mapClosure', and an analogous 'mapDirection' attribute is introduced. The new naming convention reflects attribute 'mapping' through XML trees as with

. . .. . .. . . ]]>

basically acting like

. . .. . .. . . ]]>

Webizing is optionally included in all sublanguages, eliminating the need for the entire 'ur' branch of the family of languages. However, bindatalog, bindatagroundlog and bindatagroundfact remain for the time being. See the updated modularization model for a clearer picture.

There are two kinds tags in RuleML: method-like 'role' tags and class-like 'type' tags. The two are easily distinguished by their initial letter, much like the convention used in Java: role tags start with a lower-case letter, whereas type tags start with an upper-case letter. For example, <head> is a role tag and <Ind> is a type tag.

Furthermore, type and role tags alternate in RuleML. For example, notice the alternation in this a fact asserting that John sells XMLBible to Mary:

sell John Mary XMLBible ]]>

This is known as the "striped syntax" of RuleML, and is important for compatibility with RDF. Because this is quite verbose, all role tags can also be skipped (in the spirit of Sandro Hawke's StripeSkipping), leaving a more compact form. For example, the above fact can be shortened to:

sell John Mary XMLBible ]]>

Note that there is no loss of information, because all role tags can be reconstructed based on the remaining type tags. In fact, this reconstruction (from the latter positional form to the former slotted form) can be done automatically via XSLT, as with the Normalizer in 0.88.

This gives the syntax of RuleML a lot of flexibility because it allows either a compact or expanded form (or any combination thereof), where the compact form is more user-friendly and the expanded (normalized) form is more compatible with other languages.

The most current model of the modularization of RuleML is available at http://www.ruleml.org/modularization/#Model.

The modularization used for Versions 0.7 and 0.8 was inverted in RuleML 0.85 to be more intuitive. The motivating factors behind this switch were simplicity (a single root with two distinct branches), consistency (inheritance in a single direction, for obvious super/subclass relationships) and efficiency (non-redundant implementation).

However, a limitation within XML Schema prevented this approach from being easily implemented (resulting in the unstable XSDs in 0.85), so the modularization was re-analyzed. This updated model reflects both the -- now fully-validating -- (XSD) implementation and expressiveness layering of RuleML, simultaneously capturing both the abstract and concrete.

DTDs have limited support for modularity, but it can be achieved in a roundabout way using macro-like parameter entities. In particular, the contents of an external file can be included using an externally-linked parameter entity. For example, the following includes the contents of datalog.dtd:

%datalog_include;]]>

Simple inclusion is not enough, though: overriding is also necessary. Previously, this was managed using INCLUDE/IGNORE sections: the section that declared the element which had to be changed was simply IGNOREd, then the element was re-declared.

In version 0.85, this clumsy method of overriding was replaced with a much more elegent solution wherein every element's content model was explicitly defined by a parameter entity. The rulebase label '_rbaselab' (now 'oid'), for example, became declared as follows:

]]>

Since parameter entities can overwrite one another (even across files), this content model could be easily replaced with another specified in a different DTD altogether, much like re-assigning a global variable in traditional programming languages. For example, the content model of the rulebase label '_rbaselab' is just "(ind)" in urcbindatagroundfact.dtd (as above), but was extended to permit a complex term (thus, becoming "(ind | cterm)") in hornlog.dtd:

]]>

(Note that this overriding entity must be defined before the inclusion of other files.)

A visual demonstration of this process can be found on slide 25 of the Object-Oriented RuleML: Re-Modularized and XML Schematized via Content Models presentation.

The content model-based approach to modularization also works for XML Schema, using groups (and attributeGroups) instead of parameter entities. For example (now using 0.88 syntax, where '_rbaselab' is now 'oid'),

]]>

becomes

]]>

There is no need for workarounds in XSD: <redefine> makes the specified changes and includes everything else. For example,

%include;]]>

becomes

]]>

A stable XML Schema representation of RuleML version 0.89 has been created using an approach that is consistent with that used in the (version 0.85) DTDs.

Numerous sample RuleML documents have been prepared and maintained as examples and for testing purposes. They are stored in the Examples directory. (Examples from previous versions of RuleML are also maintained, e.g. 0.88 examples).

Continuing the practice begun in 0.86, an XSLT stylesheet named 088-to-089.xslt has been created to automatically update RuleML 0.88 documents to 0.89. For instance, taking the own example from 0.88 as input, the stylesheet outputs the following: own-089.ruleml. (See a comparison of these two files generated using HTML Diff, noting that the differences are minor since 0.89 mostly builds on top of 0.88.) Additional translation output samples are found in the Upgrader directory.

An XSLT processor which may be used to perform these transformations on a whole directory at once is Saxon, using the following command:

The stylesheet has also been tested using XML Spy (version 2005 sp3).

An XSLT stylesheet has been developed (089_normalizer.xslt) for normalizing the syntax used in a given RuleML 0.89 instance, reconstructing all skipped role tags to be in a fully-expanded, normal form. For example, the compact version of the 'own' example,

buy person merchant object keep person object own person object ... keep Mary XMLBible ]]>

is normalized to the expanded version:

buy person merchant object keep person object own person object ... keep Mary XMLBible ]]>

(See a comparison of these two files generated using HTML Diff.)

Additional examples are located in the Normalizer directory.

To ensure validation stability, the 0.89 XSDs are tested (using corresponding instance documents/examples) with a selection of the most reliable validators. A summary of these validation results follows:

W3C XML Schema Validator (XSV) v 2.10-1
All examples and schemas validate perfectly except for the binary sublanguages due to a known bug introduced since 2.7-1.

Saxon v SA 8.4
All examples and schemas validate perfectly. Sample output:

Appended below is the XML Schema (version 0.89) for the Datalog sublanguage of RuleML (Appendix 1). Also appended below is a simple example rulebase that conforms to that XSD (Appendix 2), and instructions for how to validate the example against the schema (Appendix 3).

The entire family of 0.89 XSDs, specified in a modular fashion, are available here: http://www.ruleml.org/0.89/xsd. The XSD files in the main directory represent actual sublanguages of RuleML, whereas those in the modules subdirectory are elementary components included in the main XSDs. For more information, see the Modularization section.

More sample files -- each referring to the most specific XSD which validates it -- can be found at http://www.ruleml.org/0.89/exa.

Source: datalog.xsd

Source: own.ruleml