ObjectOriented RuleML for RDF: Facts, Queries, and Inferences

1
Object-OrientedRuleML for RDFFacts, Queries,
and Inferences

• Harold Boley, NRC IIT e-Business
• (with help from Said Tabet, Duncan Johnston-Watt,
  Benjamin Grosof, Bruce Spencer, Steve
  Ross-Talbot, Mike Dean, and Gerd Wagner)
• On leave from DFKI GmbH

W3C Semantic Web Architecture Meeting, Technical
Plenary and WG Meeting Event, Cambridge, MA,
USA,, 3-7 March 2003
2
Introduction

• Increased mutual RuleML-RDF(S) integration
• RDF(S) needs rules for query, inference,
  transformation
• RuleML rules need alternative syntactic encodings
  inRDF and OWL
• RuleML and OWL overlap as KRsDescription Logic
  Programs
• RuleML variables need types URIrefs to RDFS/OWL
  classes
• re A. Object-oriented RuleML is useable as an RDF
  query, inference, and transformation language
• Rules over positional (triple-like) facts 2001
• Rules over richer object-oriented facts 2003
• Queries and inferences (conjunctive) over OO
  facts(linked via named or anonymous/blank
  nodes)
• Queries and inferences over OO facts
  withbNode-embedded descriptions

Intro
3
The First RDF Graph
How to serialize this RDF graph (from MS)
Intro
4
RDF Triples as Positional Facts

• RDF triples map to positional facts, wherethe
  Property maps to a binary relation,the
  Subject to its first argument, andthe
  Object to its second argument

"http//www.w3.org/Home/Lassila has creator
Ora Lassila" ltfactgt lt_headgt ltatomgt
lt_oprgt ltrel href"http//dublincore.org/docu
ments/dces/index.shtml.rdfCreator"/gt
lt/_oprgt ltind href"http//www.w3.org/Home/Lass
ila"/gt ltindgtOra Lassilalt/indgt lt/atomgt
lt/_headgt lt/factgt
Triples Rules
_head (system) role
5
RDF RuleML Triple Roundtrip
Turn the has creator triple, as a RuleML
rulebase, again into RDF
ltrdfRDF xmlnsrdf"rdf" xmlnsruleml"ruleml"
xmlns"ruleml"gt ltrulebasegt lt_clausesgt
ltrdfSeqgt ltrdfligt ltfactgt lt_headgt
ltatomgt lt_oprgt ltrel
href"http//dublincore.org/documents/dces/index.s
html.rdfCreator"/gt lt/_oprgt lt_arggt
ltrdfSeqgt ltrdfligt ltind
href"http//www.w3.org/Home/Lassila"/gt
lt/rdfligt ltrdfligt ltind
rulemlcdata"Ora Lassila"/gt lt/rdfligt
lt/rdfSeqgt lt/_arggt lt/atomgt
lt/_headgt lt/factgt lt/rdfligt
lt/rdfSeqgt lt/_clausesgt lt/rulebasegt lt/rdfRDFgt
Triples Rules
6
RDF/RuleML Rules Over Positional Facts

• RDF/RuleML rules over positional (triple-like)
  facts provequeried triples, top-down, or derive
  new triples, bottom-up

IF "Page has creator Person" THEN
"Page was accessed by Person" ltimpgt
lt_bodygt ltatomgt lt_oprgt ltrel
href"http//dublincore.org/documents/dces/index.s
html.rdfCreator"/gt lt/_oprgt
ltvargtPagelt/vargt ltvargtPersonlt/vargt lt/atomgt
lt/_bodygt lt_headgt ltatomgt lt_oprgt ltrel
href"http//logging.org/vocabulary/xyz.rdfAccess
ed"/gt lt/_oprgt ltvargtPagelt/vargt
ltvargtPersonlt/vargt lt/atomgt lt/_headgt lt/impgt
_body (system) role
Triples Rules
_head (system) role
7
RDF Descriptions as Object-Oriented Facts

• Object-Oriented RuleML has been implemented as an
  extension of, and XSLT translator to, Positional
  RuleML
• RDF descriptions map to object-oriented facts,
  wherethe Subject maps to a relation (cf. rel.DB
  tuples),each Property maps to a role, andeach
  Object maps to its filler
• The First RDF Triple above becomes the RDF/XML
  description on the following slide
• This maps to an object-oriented RDF/RuleML fact,
  shown underneath

OO Rules
8
ltrdfRDFgt xmlnsrdf"http//www.w3.org/1999/02
/22-rdf-syntax-ns" xmlnss"http//dublincore
.org/documents/dces/index.shtml.rdf"gt
ltrdfDescription about"http//www.w3.org/Home/Las
sila"gt ltsCreatorgtOra Lassilalt/sCreatorgt
lt/rdfDescriptiongt lt/rdfRDFgt
resource ? rel literal ? ind
OO Rules
9
RDF/RuleML Rules OverObject-Oriented Facts

• RDF/RuleML rules over object-oriented facts can
  provequeried descriptions or derive new
  description-like facts

ltrulemlrulebasegt ltimpgt lt_bodygt
ltatomgt lt_oprgtltvargtPagelt/vargtlt/_oprgt
lt_r n"sCreator"gtltvargtPersonlt/vargtlt/_rgt
lt/atomgt lt/_bodygt lt_headgt ltatomgt
lt_oprgtltvargtPagelt/vargtlt/_oprgt lt_r
n"tAccessed"gtltvargtPersonlt/vargtlt/_rgt
lt/atomgt lt/_headgt lt/impgt lt/rulemlrulebasegt
IF "Page has creator Person" THEN
"Page was accessed by Person"
OO Rules
10
Bottom-Up RDF/RuleML Derivations
ofObject-Oriented Facts

• RDF/RuleML rule over above object-oriented fact
  derivesa new description in bottom-up / forward
  manner

ltrulemlrulebasegt ltfactgt lt_headgt
ltatomgt lt_oprgtltrel uriref"http//www.
w3.org/Home/Lassila"/gtlt/_oprgt lt_r
n"sCreator"gtltindgtOra Lassilalt/indgtlt/_rgt
lt/atomgt lt/_headgt lt/factgt lt/rulem
lrulebasegt
derives
OO Rules
ltfactgt lt_headgt ltatomgt
lt_oprgtltrel uriref"http//www.w3.org/Home/Lassila"
/gtlt/_oprgt lt_r n"sAccessed"gtltindgtOra
Lassilalt/indgtlt/_rgt lt/atomgt lt/_headgt
lt/factgt
new
11
Top-Down RDF/RuleML Queries OverObject-Oriented
Facts

• RDF/RuleML rule over above object-oriented fact
  provesa queried description in top-down /
  backward manner

ltrulemlquerygt lt_bodygt ltatomgt
lt_oprgtltvargtPagelt/vargtlt/_oprgt lt_r
n"tAccessed"gtltvargtPersonlt/vargtlt/_rgt
lt/atomgt lt/_bodygt lt/rulemlquerygt binds ltvargtP
agelt/vargt to ltrel uriref"http//www.w3.org/Home/
Lassila"/gt and ltvargtPersonlt/vargt to ltindgtOra
Lassilalt/indgt
WHICH "Page was accessed by
Person" ?
OO Queries
12
Intermediate Nodes for RDF Descriptions and
Object-Oriented Facts
How to serialize this RDF graph (from MS)
Named Nodes
13
Intermediate Nodes for RDF Descriptions and
Object-Oriented Facts (Contd)

• RDF descriptions and object-oriented facts can
  employnamed intermediate nodes (e.g., staff IDs
  as froma corporate relational DB used as
  URIrefs)
• The following classical two RDF/XML
  descriptionslink Oras homepage with his staff
  ID as a named intermediate node
  http//www.w3.org/staffId/85740, which gives
  further information via literals
• These map to two object-oriented RDF/RuleML facts
  linked in the same fashion, shown interleaved

Named Nodes
14
Named Nodes
15
Conjunctive RDF/RuleML Queries OverObject-Oriente
d Node-Linked Facts

• Conjunctive RDF/RuleML query of object-oriented
  facts allows a (relational-like) join over a link
  variable

ltrulemlquerygt lt_bodygt ltandgt ltatomgt
lt_oprgtltrel uriref"http//www.w3.org/Home/La
ssila"/gtlt/_oprgt lt_r n"sCreator"gtltvargtIDlt
/vargtlt/_rgt lt/atomgt ltatomgt
lt_oprgtltvargtIDlt/vargtlt/_oprgt lt_r
n"vName"gtltvargtNlt/vargtlt/_rgt lt_r
n"vEmail"gtltvargtElt/vargtlt/_rgt lt/atomgt
lt/andgt lt/_bodygt lt/rulemlquerygt binds ltvargtIDlt/
vargt to ltrel uriref"http//www.w3.org/staffId/857
40"/gt and ltvargtNlt/vargt to ltindgtOra
Lassilalt/indgt and ltvargtElt/vargt to
ltindgtlassila_at_w3.orglt/indgt
WHAT IS "/ Home/Lassila Creators
ID" AND "that IDs Name N Email
E" ?
Named Nodes
16
Blank Nodes in RDF Descriptions and
Object-Oriented Facts
How to serialize this RDF graph (from MS)
bNodes
17
Blank Nodes in RDF Descriptions and
Object-Oriented Facts (Contd)

• RDF descriptions and object-oriented facts can
  employanonymous (blank) intermediate nodes (as
  implicit in RDFs most abbreviated purely
  striped syntax)
• The following RDF/XML description embeds in
  Oras homepage a blank intermediate node, which
  gives two bits of literal information
• This maps to an object-oriented RDF/RuleML fact
  embedding an object-oriented cterm in a similar
  way, shown underneath

bNodes
18
ltrdfRDFgt ltrdfDescription about"http//www.w3.
org/Home/Lassila"gt ltsCreatorgt
ltrdfDescriptiongt ltvNamegtOra
Lassilalt/vNamegt ltvEmailgtlassila_at_w3.orglt/
vEmailgt lt/rdfDescriptiongt
lt/sCreatorgt lt/rdfDescriptiongt lt/rdfRDFgt
ltrulemlrulebasegt ltfactgt lt_headgt
ltatomgt lt_oprgtltrel uriref"http//www.w3.or
g/Home/Lassila"/gtlt/_oprgt lt_r
n"sCreator"gt ltctermgt
lt_opcgtltctor/gtlt/_opcgt lt_r
n"vName"gtltindgtOra Lassilalt/indgtlt/_rgt
lt_r n"vEmail"gtltindgtlassila_at_w3.orglt/indgtlt/_rgt
lt/ctermgt lt/_rgt lt/atomgt
lt/_headgt lt/factgt lt/rulemlrulebasegt
bNodes
19
RDF/RuleML Rules Over Object-Oriented Facts with
bNode-Embedded Descriptions

• RDF/RuleML rules over object-oriented facts can
  alsoprove queried descriptions with embedded
  descriptions

ltrulemlrulebasegt ltimpgt lt_bodygt
ltatomgt lt_oprgtltvargtPagelt/vargtlt/_oprgt
lt_r n"sCreator"gtltvargtdescriptionlt/vargtlt/_rgt
lt/atomgt lt/_bodygt lt_headgt
ltatomgt lt_oprgtltvargtPagelt/vargtlt/_oprgt
lt_r n"tAccessed"gtltvargtdescriptionlt/vargtlt/_rgt
lt/atomgt lt/_headgt lt/impgt lt/rulemlruleba
segt
bNodes
20
RDF/RuleML Queries Over Object-Oriented Facts
with bNode-Embedded Descriptions

• RDF/RuleML rule over above object-oriented fact
  provesa queried description with an embedded
  description

ltrulemlquerygt lt_bodygt ltatomgt
lt_oprgtltvargtPagelt/vargtlt/_oprgt lt_r
n"tAccessed"gtltvargtdescriptionlt/vargtlt/_rgt
lt/atomgt lt/_bodygt lt/rulemlquerygt binds ltvargtPag
elt/vargt to ltrel uriref"http//www.w3.org/Home/La
ssila"/gt and ltvargtdescriptionlt/vargt to
ltctermgt lt_opcgtltctor/gtlt/_opcgt
lt_r n"vName"gtltindgtOra Lassilalt/indgtlt/_rgt
lt_r n"vEmail"gtltindgtlassila_at_w3.orglt/in
dgtlt/_rgt lt/ctermgt
bNodes
21
Conclusions

• RDF mapped to Object-Oriented RuleMLresource ?
  rel element with a uriref attribute literal ?
  ind element (then mapped to RDF)
• Object-oriented queries can
• employ resource-linked variables in conjunctions
• invoke object-oriented rules in a backward manner
• Object-oriented rules can also be invoked in a
  forward manner to derive new OO facts, e.g. using
  CommonRules, j-DREW, Jess, or cwm
• Handle bNodes in RDF trees via embeddingin
  general RDF graphs via generated URIrefs
• Model theory can build on RuleMLs
  RDF-XML-integrating data model via F-Logic or
  TRIPLE

Concs
22
References

• Harold Boley Relationships Between Logic
  Programming and RDF, in R. Kowalczyk, S.W. Loke,
  N.E. Reed, G. Graham (Eds.), Advances in
  Artificial Intelligence, LNAI 2112,
  Springer-Verlag, 2001
• Harold Boley A Web Data Model Unifying XML and
  RDF. Draft, September 2001.
• Harold Boley The Rule Markup Language RDF-XML
  Data Model, XML Schema Hierarchy, and XSL
  Transformations, Invited Talk, INAP2001, Tokyo,
  October 2001.
• Harold Boley, Said Tabet, and Gerd Wagner
  Design Rationale of RuleML A Markup Language for
  Semantic Web Rules, Proc. SWWS'01, Stanford,
  July/August 2001.
• Andreas Eberhart, An Agent Infrastructure based
  on Semantic Web Standards, Workshop on Business
  Agents and the Semantic Web at the AI 2002,
  Calgary, Canada
• Andreas Eberhart, Automatic Generation of
  Java/SQL based Inference Engines from RDF Schema
  and RuleML, International Semantic Web Conference
  2002, Sardinia
• Benjamin Grosof Representing E-Business Rules
  for the Semantic Web Situated Courteous Logic
  Programs in RuleML, Proc. Workshop on Information
  Technologies and Systems (WITS '01), New Orleans,
  December, 2001.
• Benjamin Grosof, Mahesh D. Gandhe, and Timothy
  W. Finin SweetJess Translating DamlRuleML to
  Jess, Proc. International Workshop on Rule Markup
  Languages for Business Rules on the Semantic Web,
  Sardinia (Italy), June 2002.
• Benjamin Grosof and Terrence Poon Representing
  Agent Contracts with Exceptions using XML Rules,
  Ontologies, and Process Descriptions, Proc.
  International Workshop on Rule Markup Languages
  for Business Rules on the Semantic Web, Sardinia
  (Italy), June 2002.
• Steve Ross-Talbot, Harold Boley, and Said Tabet
  Playing by the Rules, Application Development
  Advisor 6(5), June 2002, 38-43.
• Michael Schroeder and Gerd Wagner (Eds.)
  Proceedings of the International Workshop on Rule
  Markup Languages for Business Rules on the
  Semantic Web. Sardinia, Italy, June 14, 2002.
  CEUR-WS Publication Vol-60.
• Gerd Wagner How to Design a General Rule Markup
  Language?, Invited Talk, Workshop XML
  Technologien für das Semantic Web (XSW 2002),
  Berlin, June 2002.

Refs