A FRAMEWORK FOR SEMANTICALLY RICH LEGAL
DOCUMENTS AND APPLICATIONS
John Murphy and Robert Steele
Department of Computer Science, University of Technology Sydney, PO Box 123 Sydney, 2007, Australia
Keywords:
Semantic Web Application, Web Standards, Document Annotation, RDF, OWL.
Abstract:
This paper describes the use of Semantic Web technologies for enriching legal documents within a framework
that provides a base for the development of applications to reduce problems experienced by legal practitioners
in the administration of justice. The framework builds upon the terms and concepts found within structured
information sources in the document centric legal domain. Contributions of this paper include (1) the descrip-
tion of a bottom-up ontology development approach utilizing the terms and structures within legal documents;
(2) an explanation of how this legal knowledge can be captured from existing document resources and repre-
sented using Semantic Web languages; (3) an analysis of the comparative benefits of this bottom-up approach
versus existing top-down approaches and (4) concrete examples of how to represent such legal knowledge
using OWL/ RDF. We argue such bottom-up legal ontology development grounded in the terms and concepts
found in existing legal resources will support direct and easier annotation of resources with the semantics of
legal concepts and hence provide a base for the easier development of a legal applications layer.
1 INTRODUCTION
Recent advances in semantic web technologies have
provided languages, tools and infrastructure that facil-
itate the application of these technologies to problems
in existing domains. The maturing of languages such
as OWL and RDF(S) with patterns of best practice has
enabled the application of Semantic Web technolo-
gies to domains that have the characteristics of be-
ing document centric with distributed, heterogeneous
and structured information sources. In this paper, we
report on an application of Semantic Web technol-
ogy to the legal domain and demonstrate the use of
these technologies to enrich the content of legal doc-
uments and reduce the problems faced by legal practi-
tioners dealing with complexissues using information
sources spanning multiple documents.
The domain of law shares similar characteristics
to the World Wide Web (WWW) because legal ad-
ministration is primarily document driven, with many
interrelated documents and independent authors. Le-
gal document publishing is characterized by differ-
ent types of related documents produced by differ-
ent organizations that have a mix of semi-structured
and free text parts. Unlike the WWW, the legal do-
main is underpinned by a formal legal vocabulary
that significantly reduces the problems of ambiguity.
The characteristics of formal structure, reduced am-
biguity and its distributed heterogeneous nature make
the legal domain a strong candidate for the appli-
cation of Semantic Web technologies and standards
that encourage the integration of such information
resources. Web technologies have evolved to over-
come problems of distributed heterogeneous infor-
mation sources by standardizing access to informa-
tion resources. The Semantic Web technologies of
RDF(S) and OWL havefurther enhanced the ability to
process distributed information resources by enabling
the standardization of the description of a documents
structure and the meaning of its content.
Semantic Web technologies enable a standardized
and richer description of the meaning of legal docu-
ments by annotation of the documents terms in rela-
tion to an ontology. Traditional methods of describ-
ing the meaning of documents are achieved through
external categorization systems such as indexes and
digests, that assign meaning to the whole document
using tree like structures limited to expressing the sin-
gle is
kind Of relation between parent and child con-
cepts. OWL and RDF(S) are languages that enable the
formal definition of legal concepts and the relation-
ships between those concepts using logical constructs
193
Murphy J. and Steele R. (2008).
A FRAMEWORK FOR SEMANTICALLY RICH LEGAL DOCUMENTS AND APPLICATIONS.
In Proceedings of the Fourth International Conference on Web Information Systems and Technologies, pages 193-200
DOI: 10.5220/0001523601930200
Copyright
c
SciTePress
as well as providing the means to directly annotate
those concepts within a document. Further benefits
of OWL are obtained by reasoning using the logi-
cal constructs from Description Logic that makes ex-
plicit the meaning that is implicit in the document. An
added benefit of using web standards with public le-
gal documents is that it makes the information more
accessible to the wider population and provides a plat-
form that encourages broader participation in advanc-
ing research and application development in the legal
publishing domain. Many courts and legislators are
currently publishing directly on-line or sending their
documents to commercial legal publishers or organi-
zations such as AUSTLII (AustLII, 2007) for publi-
cation on the web.
Semantic Web technologies can be applied to
solve current restrictions on searching and under-
standing information. Searching legal documents can
be extended from “bag of words” or terms to search-
ing by concepts. The development of multiple prob-
lem specific ontologies enables different perspectives
of domain knowledge to be used by applications for
searching and navigation of a set of related documents
in a more focused and better targeted manner and pro-
vides the structure for understanding those documents
from a problem specific point of view.
This paper demonstrates the use of Semantic Web
technologies to deal with some of the problems faced
by legal practitioners in the legal domain. It describes
related semantic technology research in law, explains
where our research sits with current legal ontologyde-
velopment and describes the approach we have taken
to applying the Semantic Web technologies of OWL
and RDF to the enrichment of legal documents.
2 BACKGROUND
Ontology Development. A significant part of legal
semantic research has been devoted to top-down on-
tology development to produce re-usable shared on-
tologies. The current top-down approaches to legal
ontology development have evolved from legal the-
ories such as those of (Kelsen, 1991), (Hart, 1961)
and (Bentham, 1970). These theories formed the ba-
sis of legal ontologies such as FOLaw (Valente, 1995)
that represent functional explanations of “how the law
works”. The next evolution in top-down legal ontol-
ogy development was the move toward representing
“what the law is” by capturing commonsense con-
cepts required for most applications of legal reason-
ing. LRI Core (Breuker and Hoekstra, 2004) is an
example of a core commonsense legal ontology that
reflects the roles, objects, functions and actions of the
legal system in society and are designed to facilitate
reasoning and argumentation.
Conceptual Search. Applicationsof semantic tech-
nology to conceptual searching are based on the idea
that established legal concepts can be modeled and
used for document retrieval systems that locate infor-
mation that conforms to a specific legal concept. Cur-
rent searching techniques often use syntactic or sta-
tistical analysis of terms rather than concepts or se-
mantic relations between terms. This often leads to
document retrieval that is not ranked by the expected
concept of relevance (van Noortwijk et al., 2005). For
systems to achieve the goal of conceptual retrieval
they require knowledge of legal concepts and issues
as well as knowledgeof the relationships of these con-
cepts to cases and legislation (Hafner, 1987). Our
framework establishes a base that applications can
build upon to help attain this goal.
Case Modeling. Modeling the relations between le-
gal principles is an important contribution to the task
of legal research in its own right, because some of
the associations between principals of law, legal is-
sues and facts in a sub-domain of law are not always
obvious to legal practitioners (Atkinson et al., 2005).
Recent developments in the modeling of legal argu-
mentation provide a structure for the development of
ontologies to specifically capture this important as-
pect of judicial process.
Legal Publishing on the Web. Public legal doc-
uments such as court judgements and legislation
are currently being published on websites such as
AUSTLII (AustLII, 2007). AUSTLII provides free
and open access to public legal documents on the
web using an index based on word occurrences and
databases of jurisdictions and parliaments for the pub-
lication of case law and legislation respectively. The
AUSTLII case law databases include hypertext links
to other cases, Acts of parliament and sections in
Acts.
In contrast to other frameworks that extract con-
cepts from the text using automated techniques (Bia-
gioli et al., 2005) and (Casellas et al., 2006), our work
concentrates on enriching existing legal documents
and the relationships between and within those docu-
ments focusing on those parts that provide procedural
context and summarization of the main legal points
as well as explicit external links that provide context
and supplementary information. Current approaches
to legal publishing on the web are generally restrained
to a document locator and optionally an anchor point
WEBIST 2008 - International Conference on Web Information Systems and Technologies
194
within the document. We argue there are significant
benefits to be gained in the fields of searching, docu-
ment retrieval, navigation and analysis of legal docu-
ments by enriching the documents structure, content
and links by enhancing or extending them with deeper
semantics. This we argue, will deliver more complex
and multi-faceted documents for analysis and shift the
development focus toward the concepts within docu-
ments rather than the document as a whole being the
fundamental level of conceptualization.
3 A FRAMEWORK FOR
SEMANTICALLY RICH LEGAL
DOCUMENTS AND
APPLICATIONS
The enrichment of legal documents through the ap-
plication of Semantic Web technologies begins with
the development of ontologies and the subsequent an-
notation of the existing legal documents. A bottom-
up approach to development is used taking advantage
of the structures and terminology found within legal
documents to build higher level legal concepts that
can be re-used in applications for document retrieval
and navigation, analysis and domain understanding,
document management, automated processing and le-
gal reasoning.
A good example of a structured legal document
is the common law Case Report that will be used
throughout this paper to give examples of our ontol-
ogy development approach, legal document annota-
tion process and the types of applications supported
by the framework. The Case Report document is an
important contributor to traditional paper-based legal
knowledge that contains the judgements of the court
as well as more formally structured parts contributed
by court reporters and editors as demonstrated in the
Case Report extract presented in figure 1.
Figure 1: Case Report Extract.
The extract is taken from the Head Note part of the
Case Report and shows a structured mix of judicial
case management information in the upper portion
and information concerning legal principles in the
portion marked “Held”. The Head Note in addition
to the court judgement, incorporates elements of ju-
dicial process such as hearing dates, parties and their
advocates, the presiding judge, the type of court, de-
scriptions of the nature of the case and cases cited.
Furthermore the Case Report includes more complex
parts such as the
“Catch Words”
block that is a hier-
archical categorization of key terms derived from Le-
gal Digests and a part that summarizes the important
legal issues dealt with by the case. Mixed with the
structured parts of the document are links to external
unifying sources such as common legal phrases, terms
from legal thesauri as well as links to relevant legisla-
tion.
The framework is composed of the three stages
of 1) ontology development, 2) document annotation
and 3) semantic application re-use. During the on-
tology development stage, higher level legal concepts
are evolved from the explicit legal information found
in the document text. The annotation stage involves
the semi-automated annotation of the legal text using
existing mark-up and knowledge of domain experts
such as Case Report editors whose current role in-
cludes the identification and summarization of impor-
tant legal principles. The semantic application stage
involves the selection of relevant parts of the ontolo-
gies for re-use by specific applications for purposes
such as document retrieval and management and le-
gal reasoning.
3.1 Ontology Development
The framework adopts a bottom-up approach to on-
tology development using the structures and con-
cepts contained within legal documents as the start-
ing point. This approach to ontology development
begins with the identification of concepts that are ex-
plicit in the legal documents including the names and
roles of the parties, the case citation, date of hearing,
the judges, the case and legislative citations, the ad-
vocates and references to the sections of any relevant
legislation. Figure 2 shows the conceptual model of
the Case Report document from which more abstract
concepts of law are develop to:
represent the administrative rules and processes of
the court,
encode the links between facts and legal princi-
ples,
represent the roles of the participants in litigation,
A FRAMEWORK FOR SEMANTICALLY RICH LEGAL DOCUMENTS AND APPLICATIONS
195
encode the links between the different types and
level of courts,
encode the links between legal principles, judge-
ments and previous cases incorporating the con-
cepts of Following, Not Following, Affirming, Ap-
plying and Reversing previous decisions and
accommodate the representation of a sub-domain
of law using a structure that facilitates understand-
ing of the sub-domain. For example, the sub-
domain of negligence within the laws of “Tort”,
would require a structure that encompasses the el-
ements and sub-elements of negligence, “duty of
care”, “existence and breach of the duty of care”
and “damage suffered as a result of the breach”.
Figure 2: Case Report Concepts.
Bottom-up ontology development starts with the se-
lection and definition of of the most specific concepts
at the bottom of the hierarchy. The design advances
toward more abstract concepts by grouping the lower
concepts into more general concepts and by the ap-
plication of rules to the lower concepts for the con-
struction of higher level concepts. This bottom-up ap-
proach yields the benefits of accessibility and famil-
iarity to the domain experts that annotate legal docu-
ments.
An example of developing higher level legal con-
cepts from specific concepts explicit in the document
is the Considered Case concept that is implied by the
citation link identified in the Case Report extract in
figure 1 as “AAA v BBB (1998) 23 FAM LR 716, af-
firmed”. This is the citation of a case that the judge
took into consideration when deciding the
KR v KR
case specifically Affirming the previous decision. The
Considered Case is a higher level legal concept and
judicial mechanism that supports the related concept
of Case Authority in common law that encompasses
both the adoption or rejection of legal precedent and
defined by the (NYSUC, 2007) as a:
previously adjudged action or decision on
same or similar point, serving as a rule or ex-
ample for present guidance.
Case Authority can be interpreted as the application of
cumulative weightings of importance to a case over
time using the decisions of subsequent judgements
that have considered the case and Followed, Not Fol-
lowed, Reversed or Applied the case rulings.
A number of benefits can be obtained using the
framework that include:
Accessibility by legal knowledge experts and le-
gal practitioners because the approach builds upon
the direct annotation of the original document
source and the legal domain is a knowledge in-
tensive domain that stores its knowledge in doc-
uments that can be marked-up using XML. This
direct access to the source by means of a readily
available technology gives non-technical person-
nel easier access to the source and enables broader
participation in the capture of knowledge from the
source documents.
Familiarity of concepts used by domain experts
because top-down approaches require knowledge
of higher level concepts defined by external par-
ties that are often highly abstract with subtle dis-
tinction that are not easy to understand and are
difficult to use as a starting point for development.
Better management of the development process
by encouraging partitioning of concepts along fa-
miliar domain and task specific lines.
Our research project is developing ontologies to
represent the knowledge of Case Report structures
and case reporting rules with a mixture of complex
and simple relations between legal concepts. The on-
tologies are designed to allow extension for the cap-
ture of complex legal principles and assist with tasks
such as indexing, extraction and summarization of le-
gal issues and the provision of an entry point for do-
main understanding and conceptual search for legal
practitioners.
3.2 Annotation
After the required legal concepts are defined in on-
tologies, the lower level concepts are identified in the
documents and annotated. This process can be semi-
automated using existing mark-up to assist the iden-
tification of terms. Lower level ontology concepts
can often be mapped directly to explicitly marked-
up terms in the text whilst other mark-up that iden-
tifies document structures can be used to locate terms.
Our approach to the ontology development yields the
WEBIST 2008 - International Conference on Web Information Systems and Technologies
196
benefit of starting with the most specific concepts that
are explicit within the legal document or can be easily
identified with human assistance.
A feature of legal documents that makes them
more conducive to a bottom-up approach is that much
of the publicly available legal documentation such
as Case Reports and legislation already have their
structure and many of the lower level explicit con-
cepts marked-up in XML or SGML. Many jurisdic-
tions have marked-up their court documents at the ini-
tial point of production by the the judge or judges as-
sociate and their work has been further expanded by
Case Reporters who add value to the judgements by
making explicit many concepts in the judgement doc-
ument that are implicit. A benefit of the bottom-up
approach is that it only requires the mark-up of the
lower level legal concepts which facilitates the anno-
tation of this large body of existing legal documenta-
tion through automatic or semi-automatic transforma-
tions of the existing XML or SGML.
Further benefits of direct annotation of legal doc-
uments are that it
removes the impedance of external referencing of
sources found in other types of knowledge stores
such as databases or technologies acting external
to the source such as Xpointer.
relieves the complications associated with au-
tomatic identification of concepts using Natural
Language Processing by relying on the interven-
tion of human legal experts such as drafters of leg-
islation and Case Report editors.
A final benefit is the industry and research commu-
nity backing that is derived from the use of widely ac-
cepted standards based mark-up languages. This open
community backing ensures the availability of qual-
ity editing tools and repository and reasoning servers
along with a large body of published knowledgein the
Semantic Web technologies field.
An example of the direct annotation of terms in a
legal document is shown in figure 3.
Figure 3 demonstrates the annotation of two of the
components of the example citation AAA v BBB
(1998) 23 FAM LR 716, affirmed” given previouslyin
the Case Report extract in figure 1. This is achieved
using the XML representation of OWL to mark-up
AAA v BBB” as an instance of the Considered Case
concept that is implicit in the citation and to de-
fine “AAA v BBB” as having a has considered value
property of “affirmed” thereby defining AAA v
BBB” a previously considered case that has been af-
firmed.
Figure 3: Case Report Mark-up.
3.3 Semantic Applications
After the conceptual infrastructure of the ontologies
and annotated documents are created, the applica-
tions layer of the framework can be used to de-
velop applications for purposes including document
retrieval, navigation and domain understanding, doc-
ument management, legal process automation and le-
gal reasoning.
3.3.1 Document Retrieval, Navigation and
Domain Understanding
Understanding legal concepts and arguments is im-
portant in the legal domain and dealing in concepts
rather than documents enhances the access and un-
derstanding of legal knowledge particularly when the
understanding of a legal concept requires information
spanning multiple documents. In cases where docu-
ments are made available through web pages, HTML
provides simple links that have led to fields of knowl-
edge discovery built from the statistical analysis of
links into and out of a page, web site or domain. The
introduction of conceptual links brings greater com-
plexities to the link relationships and requires a Se-
mantic Web language to define and express the nec-
essary information. Links can be enriched using RDF
to hold more properties than the traditional resource
locator and it is possible for links to be multifaceted
and allow for reasoning over legal concepts by using
OWL to add higher level concepts. Potential bene-
fits of this type of annotation are the development of
applications that bring about
a reduction in time searching legal documents
a reduction in the amount of non relevant material
the legal practitioner must process and
A FRAMEWORK FOR SEMANTICALLY RICH LEGAL DOCUMENTS AND APPLICATIONS
197
a reduction of the amount of instruction a legal
practitioner must give to a computer to retrievethe
required information.
There is a view held that future search engines will
rely less on ranking of relevance of complete docu-
ments and behave more like information management
tools (Bontcheva et al., 2006, p143). These engines
will analyse and summarize the content of documents
rather than listing the documents by relevance and
leaving the determination of the documents meaning
to the user. Information will be defined by meaning
and not statistical coincidence of terms and search-
ing will not be for documents, but for ideas or con-
cepts within documents or across documents. Legal
documents such as a Case Reports or legislation have
structures that deal with multiple concepts of law. The
evolution and refinement of legal concepts across ju-
dicial processes such as court appeals and reapplica-
tion of the law to new facts, reinforces the importance
of understanding concepts across documents.
3.3.2 Document Management and Automated
Processing
A significant problem in the legal publishing domain
is changes to documents such as legislation and its
contained law, and the changes to legal concepts as
case law evolves. This problem is separate to the sub-
stantive legal issues dealt with by legal practitioners
but is vital to finding and understanding the applica-
ble laws. The repeal of legislation, the affirmation
or reversing of court decisions and the jurisdictional
boundaries and authority of courts are all examples
of metadata contained in legal documents that can be
enriched to provide better ways to manage legal doc-
uments to assist in better understanding the applica-
tion of the law. Document management of court cases
and legislation is particularly important for the main-
tenance of the state of the law because laws should
not be retrospectively applied by courts and therefore
it is essential to know the state of the law at the point
in time of the occurrence of some legal event.
Legislation has a number of stages that it must pass
before it becomes law. Courts also have a number
of steps that a document must pass through in or-
der to comply with court procedures. The annotation
of legal documents provides the information required
for systems to track the progress of the production of
these legal documents.
3.3.3 Legal Reasoning
Much effort has been invested by the AI community
to develop decision support systems built on differ-
ent forms of legal reasoning including, case based,
rule based and Description Logics (DL) based rea-
soning. Our approach limits the use of reasoning to
the evolution of higher level legal constructs. The
OWL language provides for reasoning using Descrip-
tion Logics and more recently, a constrained horn
clause language called the Semantic Web Rule Lan-
guage (SWRL). SWRL is a formal submission to the
W3C (Horrocks et al., 2004) for the introduction of
horn clause statements that can be used for the ex-
pression of rules such as those found in legislative
norms. The use of horn clauses to express legal norms
is demonstrated by (McCarty, 1989) in his Language
for Legal Discourse and also discussed in (Visser and
Bench-Capon, 1998) and can be used to express some
norms that can not be easily expressed using DL.
The legal concept of affirmed on Appeal provides a
good example of the difficulty to assert in OWL DL
that a case Appealed and Affirmed in the same hearing
is a case Affirmed on Appeal. By contrast, the same
concept can easily be expressed in SWRL as
Af firmedOnAppeal(x) A f firmedIn(x, y)
AppealedIn(x, z) sameHearingAs(y, z)
The reasoning capabilities of OWL DL can be used
to classify new concepts and can therefore be used
to create hypothetical scenarios that can be con-
structed using OWL class constructors and classified
for querying purposes.
A further benefit of using the OWL language is that
the reasoning capabilities it provides eases the anno-
tation process in circumstances where a simple term
in the legal document is to be further developed as a
more complex legal concept. This benefit is demon-
strated using the the previous case citation example
shown in the Case Report extract in figure 1. The
term Affirmed is a member of a collection of attributes
for determining the impact a Considered Case has on
a previous court decision along with the terms Fol-
lowed, Not Followed, Applied and Reversed. At face
value, these terms are merely a list of discriminating
attributes for a Considered Case and would be created
as instances of an OWL class and be easily marked-
up in the legal document. If however the Affirmed at-
tribute is to be expanded into a richer concept, it must
be created as an OWL class and therefore can not be
annotated directly in the document without also cre-
ating a specific
AAA v BBB
instance of the new Af-
firmed class.
Figure 4 demonstrates how the OWL language en-
ables the individual instance
AAA v BBB
to be de-
fined as a type of anonymous class having some value
WEBIST 2008 - International Conference on Web Information Systems and Technologies
198
Figure 4: Case Citation Representation.
from Affirmed Case for its has considered value
property.
This enables a reasoner to infer that
AAA v BBB
is an
Affirmed Case
without having to create a par-
ticular instance of Affirmed Case for
AAA v BBB’s
has considered value property. The term Affirmed
can now be expanded as a legal concept to incorpo-
rate informationabout the jurisdictional hierarchy and
relative superiority of the court that heard the case or
could include a temporal dimension that indicates the
changing authority of a case over time. This exam-
ple demonstrates the use of new Semantic Web tech-
nologies and their associated patterns of best practice
(Rector and Welty, 2005) to enrich the concepts found
in legal documents in ways previously unavailable.
4 CONCLUSIONS
The Semantic Web technologies of RDF and OWL
enable existing structured and semi-structured legal
documents to be enriched to formally represent the
complex and interrelated concepts found within these
documents. We have described a framework that em-
ploys a bottom-up approach to developing legal on-
tologies grounded in the terms and structures of the
documents and re-using much of the existing mark-up
found in todays legal documents as well as the skills
of domain experts such as editors, case reporters and
legislative drafters. The framework also provides for
the direct annotation of these documents by the do-
main experts and the selective use of the the ontolo-
gies in semantic applications. We have also discussed
the benefits of this approach and give concrete ex-
amples in OWL and RDF. It is considered that such
an approach will facilitate more direct and easier le-
gal document annotation, as the ontology concepts are
groundedin the existing terms found in the documents
themselves.
REFERENCES
Atkinson, K., Bench-Capon, T., and McBurney, P. (2005).
Arguing about cases as practical reasoning. In ICAIL
2005: Proceedings of the 10th international confer-
ence on Artificial intelligence and law, pages 35–44.
ACM Press, New York, NY, USA.
AustLII (2007). Australasian legal information institute.
http://www.austlii.edu.au/.
Bentham, J. (1970). An Introduction to the Principles of
Morals and Legislation. Athlone Press, London.
Biagioli, C., Francesconi, E., Passerini, A., Montemagni,
S., and Soria, C. (2005). Automatic semantics ex-
traction in law documents. In ICAIL 05: Proceedings
of the 10th international conference on Artificial in-
telligence and law, pages 133–140. ACM Press, New
York, NY, USA.
Bontcheva, K., Davies, J., Duke, A., Glover, T., Kings, N.,
and Thurlow, I. (2006). Semantic Web Technologies:
trends and research in ontology-based systems, chap-
ter 8, pages 139–169. John Wiley & Sons, Ltd.
Breuker, J. and Hoekstra, R. (2004). Epistemology and on-
tology in core ontologies: Folawand lri-core, two core
ontologies for law. In Proceedings of EKAW Workshop
on Core ontologies. CEUR.
Casellas, N., Jakulin, A., Vallb, J.-J., and Casanovas, P.
(2006). Acquiring an ontology from the text. In Ali,
M. and Dapoigny, R., editors, Advances in Applied Ar-
tificial Intelligence, pages 1000–1013. Springer.
Hafner, C. D. (1987). Conceptual organization of case law
knowledge bases. In ICAIL ’87: Proceedings of the
1st international conference on Artificial intelligence
and law, pages 35–42, New York, NY, USA. ACM
Press.
Hart, H. (1961). The Concept of Law. Clarendon Press,
Oxford.
Horrocks, I., Patel-Schneider, P. F., Boley, H., Tabet, S., and
Dean, B. G. M. (2004). SWRL: A Semantic Web Rule
Language Combining OWL and RuleML. Technical
report, W3C.
Kelsen, H. (1991). General Theory of Norms. Clarendon
Press, Oxford.
McCarty, L. T. (1989). A Language for Legal Discourse.
In Proceedings of the Second International Confer-
ence on Artificial Intelligence and Law, pages 180–
189, Vancouver, Canada.
NYSUC (2007). New York State Unified Court System,
law libraries, glossary of legal terms.
Rector, A. and Welty, C. (2005). Simple part-whole re-
lations in OWL Ontologies. Best practices, W3C,
http://www.w3.org/2001/sw/BestPractices/OEP/ Sim-
plePartWhole/. Retrieved August 2007.
A FRAMEWORK FOR SEMANTICALLY RICH LEGAL DOCUMENTS AND APPLICATIONS
199
Valente, A. (1995). Legal knowledge engineering: A mod-
elling approach, volume 30 of Frontiers in artificial
intelligence and applications. IOS Press.
van Noortwijk, K., Visser, J., and Mulder, R. V. D. (2005).
Re-usable retrieval concepts for the classification of
legal documents. In Proceedings of the 10th inter-
national conference on Artificial intelligence and law,
pages 252–253. ACM Press, Bologna, Italy.
Visser, P. R. S. and Bench-Capon, T. J. M. (1998). A
comparison of four ontologies for the design of legal
knowledge systems. Artificial Intelligence and Law,
6(1):27–57.
WEBIST 2008 - International Conference on Web Information Systems and Technologies
200