| AITEC Contract Research Projects in FY1997 : Software |
| Principal Investigator : | Dr. Katsumi Nitta, Prof. |
| Tokyo Institute of Technology |
Shaky and Rbrs are legal knowledge representation languages. Shaky was
designed from viewpoint of description of legal knowledge and Rbrs was
designed from viewpoint of nonmonotonic reasoning.
(1)Shaky 2.0
Legal Reasoning System Shaky v2.0 provides the denotation of
situation, the class of term and predicate and the two features of
proposition (or predicate) for describing legal knowledge. Shaky can
distinguish between the proposition declared to be an event and the
one declared to be a property in a situation. The property is
continuous on an interval but the event not. In the system, the term
and the predicate that can represent the class as a set construct each
class-hierarchy of them. On the class-hierarchies, the system realizes
the unification of terms or predicates based on the two features of
proposition.
Shaky can be used not only the problems of legal reasoning but also
general ones. Sample programs show that this system can describe them
and derive inferences adapted from human intuition.
(2)Rbrs 1.0
RBRS is a software which helps users to make arguments for a given
claim. It is composed of an inference engine of a prioritized
Extended Logic Programming Language, and a graphical user interfaces
which shows relations between arguments. The feature of RBRS is as
follows.
(i) A program of the prioritized logic programming language is a set
of default rules. For a given claim, the inference engine makes
not only arguments which support it, but counterarguments which
attacks arguments. The conflict between an argument and its
counterargument is resolved by calculationg priority between rules
which were used in arguments. The priority between rules is also
a claim which is supported by an argument (meta-argument).
For a given claim, by considering all related arguments and meta-
arguments, the inference engine decides the strength of the claim.
This function is convenient to develop a legal reasoning system
because making meta-arguments is one of important processes of
legal disputation.
(ii) As the inference engine makes several arguments and meta-arguments.
As the relation among them tends to become very compricated, RBRS
shows them in the hierarchical manner.
Users of the system can see the detail of each argument by
clicking the corresponding box in the display.
(1)Shaky 2.0 SunOS 4.1.4 SICStus Prolog ver.3 (2)Rbrs 1.0 SICStus Prolog Ver.3 and Tcl7.6/Tk4.2
(1)Shaky
+- Readme_J(in Japanese)
+- Readme_E(in English)
+- Install_J(in Japanese)
+- USERMAN_J(in Japanese)
+- [src]
+- Makefile
+- io.pl
+- system.pl
+- la.pl
+- main.pl
+- transfer.pl
+- db.pl
+- makefile.pl
+- grammar.pl
+- utility.pl
+- [sample]
+- sample1
+- sample2
+- sample3
+- sample4
(2)Rbrs
RBRS consists of following files.
INSTALL install manual
USERMAN users' manual
comparePRule.pl inferring rule priority
compareRule.pl inferring rule priority
debate_Dic a sample dictionary
debate_Rule a sample rule base
findAllArg.pl making all arguments
findIssue.pl finding issue points
interface.tcl Tcl/Tk module
makeArg.pl making arguments
makeCounterArg.pl making all counterarguments
makaclause.pl compiling rules
makelower.pl compiling a dictionary
operator.pl declaration of operators
priority.pl calculating prioriity
rbrs shell script for starting system
startscript.pl loading initial module
tcltkinterface.pl user interface
unify.pl unification of Psi-term
cbr.pl case based reasoner
www-admin@icot.or.jp