[Paper Review] An Argumentation-Based Legal Reasoning Approach for DL-Ontology

Zhe Yu, Yiwei Lu|arXiv (Cornell University)|Sep 7, 2022

Semantic Web and Ontologies4 citations

TL;DR

This paper proposes an argumentation-based legal reasoning framework, L-ASPIC+, that integrates description logics (DL) with formal argumentation to handle inconsistency in legal ontologies, particularly for autonomous vehicle design. By translating DL-based legal ontologies into argumentation theories and applying preference-based reasoning, the approach enables consistent, explainable legal conclusions even under conflicting norms.

ABSTRACT

Ontology is a popular method for knowledge representation in different domains, including the legal domain, and description logics (DL) is commonly used as its description language. To handle reasoning based on inconsistent DL-based legal ontologies, the current paper presents a structured argumentation framework particularly for reasoning in legal contexts on the basis of ASPIC+, and translates the legal ontology into formulas and rules of an argumentation theory. With a particular focus on the design of autonomous vehicles from the perspective of legal AI, we show that using this combined theory of formal argumentation and DL-based legal ontology, acceptable assertions can be obtained based on inconsistent ontologies, and the traditional reasoning tasks of DL ontologies can also be accomplished. In addition, a formal definition of explanations for the result of reasoning is presented.

Motivation & Objective

• Address the challenge of reasoning with inconsistent legal ontologies in AI systems, particularly in high-stakes domains like autonomous vehicle design.
• Overcome the limitations of traditional DL reasoning, which struggles with inconsistency and lacks user-friendly explanations.
• Develop a formal, explainable reasoning system that supports normative legal reasoning by integrating legal principles and argument preferences.
• Enable engineers to reason about legal obligations in complex scenarios—such as drunk passengers after accidents—using structured, justifiable conclusions.
• Provide formal explanations for why certain legal conclusions are accepted or rejected, grounded in legal rules and principles.

Proposed method

• Extend the ASPIC+ framework into L-ASPIC by incorporating legal principles and normative rules to support defeasible reasoning in legal contexts.
• Translate DL-based legal ontologies into argumentation theory by mapping concepts, roles, and axioms into logical formulas and inference rules.
• Define argument construction using premises from the ontology and rules from the legal normative system, with strict and defeasible rules distinguished.
• Introduce preference relations between arguments based on legal principles, using a priority function to resolve conflicts.
• Apply argumentation semantics (e.g., credulous/sceptical) to determine acceptable conclusions from inconsistent rule sets.
• Formalize explanations as sets of premises, rules, and preference orders that justify the acceptance or rejection of a legal assertion.

Experimental results

Research questions

• RQ1 How can inconsistent legal ontologies be reasoned with effectively in AI systems, especially in safety-critical domains like autonomous vehicles?
• RQ2 How can formal argumentation be integrated with DL-based ontologies to preserve reasoning capabilities while handling normative conflicts?
• RQ3 What mechanisms ensure that legal conclusions are explainable and justifiable to engineers and regulators?
• RQ4 How can legal principles be formally encoded to guide argument preferences and resolve conflicts between competing norms?
• RQ5 What is the formal structure of an explanation that clarifies why a specific legal conclusion is accepted or rejected?

Key findings

• The L-ASPIC+ framework successfully handles reasoning over inconsistent DL-based legal ontologies by leveraging argumentation to resolve conflicts between legal norms.
• The approach supports traditional DL reasoning tasks such as concept instantiation and instance retrieval through the combined argumentation-theory framework.
• Explanations for legal conclusions are formally defined as the union of premises, rules, and preference orders, enabling traceable justification of reasoning outcomes.
• In the autonomous vehicle case study, the system determines whether a driverless car should require a drunk passenger to assist an injured person, based on normative priorities.
• The framework distinguishes between core and auxiliary functions in AV design, using legal principles to prioritize obligations like emergency aid over route efficiency.
• For the assertion 'LeaveCar(PS1)', the explanation includes premises like 'NeedEmergencyAid(Injury1)' and rules 'r8', 'r9', with preference 'p2 < p1' justifying acceptance.