Modeling Administrative Discretion Using Goal-Directed Answer Set Programming

tags

Rules as Code

Notes

none of the existing proposals are able to represent the ambiguity and/or administrative discretion present in contracts and/or applicable legislation

NOTER_PAGE: (1 . 0.7077380952380953)

a framework, called s(LAW), that allows for modeling legal rules involving ambiguity, and supports reasoning and inferring conclusions based on them

NOTER_PAGE: (2 . 0.24523809523809526)

s(CASP)

NOTER_PAGE: (2 . 0.5767857142857143)

This work makes two main contributions: (i) a set of patterns to translate legal rules into ASP, and natural language patterns to generate readable justifications; (ii) a framework to model, reason, and justify conclusions based on the evidence provided by the user and the applicable legislation, representing ambiguity, discretion and/or incomplete information (key concepts in legal cases).

NOTER_PAGE: (3 . 0.3285714285714286)

a legal article is a default rule subject to possible exceptions

NOTER_PAGE: (3 . 0.6702380952380953)

discretion to act introduces different possible interpretations of the law

NOTER_PAGE: (5 . 0.3529761904761905)

Implementations based on Prolog compute a single, canonical model, and therefore, bypass this non determinism

NOTER_PAGE: (5 . 0.38333333333333336)

discretion to act can be considered as a ground or an exception

NOTER_PAGE: (5 . 0.4154761904761905)

desirable behavior should capture the absence of information by generating different models depending on the relevant information

NOTER_PAGE: (5 . 0.6232142857142857)

The main advantage of s(LAW) is its ability to generate justifications not only for positive but also for negative information.

NOTER_PAGE: (8 . 0.5065476190476191)

During the grounding phase, links between variables are lost and therefore an explanation framework for these systems must face many challenges

NOTER_PAGE: (9 . 0.21250000000000002)

systems that follow a top-down execution can trace which rules have been used to obtain the answers more easily

NOTER_PAGE: (9 . 0.3642857142857143)