® InfoJur.ccj.ufsc.br
Instituut voor Taal- en Kennistechnologie
Institute for Language Technology and Artificial Intelligence
System Descriptions
John Zelesnikov
The IKBALS Project (Intelligent Legal Knowledge BAsed Systems) at the Database Research
Laboratory, Applied Computing Research Institute, La Trobe University, Victoria,
Australia, has as its aim the development of intelligent legal support tools.
The original prototype IKBALS I was a hybrid/object oriented rule based system built in
the Goldworks II object oriented environment. Its descendants, IKBALS II and IKBALS III
provide for integrated rule based reasoning, case based reasoning, and information
retrieval by using intelligent cooperating information systems. They were both build using
the NExpert Object expert system shell. IKBALS I and IKBALS II provided advice on the
Accident Compensation Act (Victoria, Australia, 1985). IKBALS III provides advice on the
Credit Act (Victoria, Australia, 1984).
The main features of the integrated IKBALS III system are:
- The use of cooperating agents;
- The use of an application programming interface to act as a bridge between agents in the
IKBALS III system and the artificial intelligence kernel in the Knowledge Base;
- The use of a customised induction algorithm that generated the indices into the case
base;
- The use of background information to supplement the induction process;
- A method for converting the decision tree produced by the induction algorithm into
quantitative knowledge based rules;
- The heuristics used to justify explanations;
- The query facility that enables users to investigate the relationship between cases and
arguments in the system.
In conjunction with a firm of solicitors, the IKBALS project has developed the Credit
Act Advisory System (CAAS). CAAS is a strictly rule based legal expert system which
advices in relation to a small part of the Credit Act 1984 (Vic.). It is essentially the
rule based component of IKBALS III. CAAS does not argue directly with either statutes or
precedents. Instead, it is a production rule system where the production rules forming the
knowledge base of CAAS are heuristic rules supplied by experts from Allan Moore & Co.
It has been prototyped using NExpert Object and then compiled into C++ under Windows 3.1.
In developing CAAS, the IKBALS project has developed a framework for building commercial
legal expert systems using C++, rather than the more expensive expert system shells.
References
- Vossos, G., Zeleznikow, J. and Hunter, D.
- (1993), `Building Intelligent Litigation Support Tools through the integration of Rule
Based and Case Based Reasoning', Law, Computers and Artificial Intelligence 2
(1), pp. 77--93.
- Vossos, G., Zeleznikow, J., Moore, A. and Hunter, D.
- (1993), `The Credit Act Advisory System (CAAS): Conversion from an expert system
prototype to a C++ commercial system', Proceedings of the Fourth International
Conference on Artificial Intelligence and Law, ACM Press, pp. 180--183.
- Zeleznikow, J.
- (1991), `Building Intelligent Legal Tools - The IKBALS Project', Journal of Law and
Information Science 2 (2), pp. 165--184.
- Zeleznikow, J., Vossos, G. and Hunter, D.
- `The IKBALS Project: Multimodal reasoning in Legal Knowledge Based Systems' to appear in
AI and Law: An International Journal.
John Zeleznikov
SPLIT-UP is a hybrid rule based expert system developed at the Database Research
Laboratory, Applied Computing Research Institute, La Trobe University. It was commissioned
by a judge of the Family Court of Australia.
SPLIT-UP uses textbooks, heuristics, expert advice and cases to model that part of the
Family Law Act 1975 (Australia) which deals with property division. When settling property
cases, judges need to determine which assets she has the power to distribute (known as the
Common Pool) and then decide what percentage of the Common Pool each partner receives. The
module for determining the Common Pool is rule-based.
Because the distribution of assets in the Common Pool, is discretionary, the developers
opted to use neural networks to determine the appropriate percentage splits. The original
rule based system to determine the contents of the Common Pool was written in the expert
system shell VP-Expert. The final system is written in C++ and has machine learning and
neural network enhancements. It focuses upon an explanation for its advice.
References
- Stranieri, A. and Zeleznikow, J.
- (1992), `SPLIT-UP Expert System to determine Spousal Property distribution on Litigation
in the Family Court of Australia', in Adams, A. and Sterling, L. (eds.), Proceedings
of AI'92, World Scientific, 1992, pp. 51--56.
- Stranieri, A., Hunter, D. and Zeleznikow, J.
- (1994), `Open texture and bound: Dimensions for Artificial Intelligence and
Jurisprudence', submitted to AI and Law: An International Journal.
Egon Verharen and Wim Voermans
In 1991 the Dutch Ministry of Justice was interested in investigating the possibilities
of current computer science applications for supporting the activities of legal civil
servants during the preparation process of regulation and legislation. Secondly, the
Ministry wanted to look into the possibility of automatic disclosure of the 'Aanwijzingen
voor de Regelgeving' (Recommendations for Regulation: the rules and guidelines civil
servants have to obey when making regulation) and the advantages of this over the ``paper
situation''.
Wim Voermans of the Faculty of Law, and Egon Verharen of ITK at Tilburg University
found that hypertext facilities were extremely suitable for not only the disclosure of the
Recommendations in several ways, but also for structuring the knowledge included in the
preparation activities for legislation. As a result of the research LEDA I (LEgislative
Design and Advisory system), an experimental model for a legislative advisory and design
system, was developed. It is a prototype of a semi-intelligent authoring tool that, using
hypertext and language technology techniques, offers information of importance and
interest in the preparation and design of legislation to legal draftsmen and legislators.
Furthermore it offers active methodological support in the actual design and layout of
legal documents and access to on-line databases and documentary information servers.
The first prototype was developed using ToolBook. It works on any 386sx or better,
windows 3.1 and has already been operational for testing purposes at the Ministry since
1993. Right now we are working on LEDA II, a new prototype in C++, that will be
operational this summer. Please contact us if you are interested in the LEDA system and
would like to receive one of the limited number of a demo-version of LEDA I. Also, if you
are interested in (any paper of) the LEDA documentation, please contact us on how to get
hold of a copy (if public and available).
The LEDA Papers
Reports
- W. Voermans
- Computerhulp bij wetgeving, rapport van een verkennende studie inzake de
mogelijkheden van vormen van computerondersteuning bij de uitvoering van wetgevingstaken
en -activiteiten, uitgevoerd met subsidie van het Ministerie van Justitie, Tilburg
januari 1992, X + 240 pp. (in Dutch)
- W. Voermans, L. Matthijssen
- Computerhulp bij wetgeving II; Onderzoeksrapport Faculteit der Rechtsgeleerdheid/ITK,
Katholieke Universiteit Brabant, Tilburg december 1993. (in Dutch)
- L.J. Matthijssen
- Documentstandaard voor een wetgevingsontwerp- en adviessysteem,
Onderzoeksrapport Instituut voor Taal- en Kennistechnologie, Katholieke Universiteit
Brabant, Tilburg maart 1993. (in Dutch)
- M. Neut
- LEDA-documenten omzetten van ODA naar SGML en vice versa; onderzoeksrapport
Instituut voor Taal- en Kennistechnologie, Katholieke Universiteit Brabant, Tilburg,
januari 1994. (in Dutch)
Papers
- W. Voermans
- `Het teken, dat u bewaart en duidt tot in verste verandering', in: H.A.M. Backx e.a., Recht
doen door wetgeving, Opstellen over wetgevingsvraagstukken aangeboden aan mr. E.M.H.
Hirsch Ballin, W.E.J. Tjeenk Willink, ISBN 90 271 3173 2, Zwolle 1990, p. 323--336 (in
Dutch)
- W. Voermans
- `Ontwerpen van wetgeving met behulp van de computer', in: RegelMaat,
kwartaalblad voor wetgevingsvraagstukken, jg. 1990, afl. 1, Koninklijke Vermande b.v.,
ISSN 0920 055X, Lelystad 1990, p. 17--22 (in Dutch)
- W. Voermans
- `Computer-aided Legislative Design: Worth while the Effort?', in: C. van Noortwijk,
A.H.J. Schmidt, R.G.F. Winkels (eds.), Legal knowledge based systems; Aims for
research and development, Koninklijke Vermande b.v., ISBN 90 6040 948 5, Lelystad
1991, p. 87--97
- E. Verharen, W. Voermans
- `Ontwerp en implementatie van LEDA', hypertextworkshop bijdrage Eindhoven 26 maart 1992,
TU Eindhoven 1992
- E. Verharen, M. Fridael, W. Voermans
- `Ontwikkeling van en hypertexttoepassingen binnen LEDA', in: H. Weigand, H. Paaijmans
(red.), Workshop Artificial Intelligence en Information Retrieval, ITK, ISBN 90
74029 04 3, Tilburg 1992, p. 44--53 (in Dutch)
- F. Maes, E.J. van der Linden
- `De bruikbaarheid van online-hulpstructuren bij hypertexttoepassingen', in: H. Weigand,
H. Paaijmans (red.), Workshop Artificial Intelligence en Information Retrieval,
ITK, ISBN 90 74029 04 3,Tilburg 1992, p.54--61 (in Dutch)
- W. Voermans, A.H. de Wild, F.D.H.M. Maas en J. Nouwt
- `Are Legislators beyond Help from Legal Computer-Science?', in: AIgg Kennisgeving,
jg. 5, nr. 3, september 1992, p. 2--11
- F. Maes, S. Goutier, E.J. van der Linden
- `Online reading and offline tradition: adapting online help facilities to offline
reading strategies', in: Proceedings of SIGDOC'92: 10th aAnnual International
Conference on Documentation, Ottawa, Canada, October 13--16 1992, ACM, 1992
- E. Verharen, M. Fridael, W. Voermans
- `Experimenteel model LEDA: kennisgebaseerd gebruik van hypertexttechniek in een
wetgevingsontwerp- en adviessysteem', in: B.R. van der Spek, R. de Hoog (eds.), Proceedings
Kennistechnologie '92; methoden, technieken en gereedschappen, ISBN 90 71694 17 8,
Rijswijk 1992, p. 329--340 (in Dutch)
- E. Verharen, M. Fridael, W. Voermans
- `Hypertext-toepassingen binnen een wetgevingsontwerp- en -adviessysteem', in: N.J.I.
Mars (red.), Informatiewetenschap 1992 (wetenschappelijke bijdragen aan de tweede
Stinfonconferentie d.d. 18 december 1992), pp. 89--109 (in Dutch)
- E. Verharen, W. Voermans
- `Leda: a semi-intelligent drafting support systemen', in: J.G.J. Wassink, J. Svenson
(eds.), Intelligent tools for designing legislation and computer-supported comparison
of law, Proceedings of the 1993 Jurix-conference TU Twente, Enschede, Lelystad 1993.
- E. Verharen, M.R. Fridael, W. Voermans, L.J. Matthijssen
- `Leda and Obw: Two semi-intelligent legislative drafting support systems', bijdrage aan
de conferentie Towards a Global Expert System in Law, Florence 1--3 december 1993. (To
appear).
© Arthur van Horck