Back to search

National Technical University of Athens (NTUA)

Argumentation and rule-based logic in mathematical proving and legal artificial intelligence applications

Abstract

dc:description

Proving is the validation of conclusions by application of logical arguments and rules to assumptions. In mathematics, an assertion is not accepted as true or valid unless it is accompanied by a proof. However, proving processes do not exist only in mathematics, but almost everywhere — in the physical sciences, in computer science, in legal and ethical argumentation, in philosophy, and so on. During the proving process, a dialogue between agents is required to clarify obscure inference steps, fill gaps, or reveal implicit assumptions in a purported proof. Hence, argumentation is an integral component of the discovery process of proofs in general but also — more specifically — in mathematical proofs. The first part of this thesis presents how logic-based argumentation theories can be applied to describe specific features in the development of proof-events, highlighting the relation between formal proof, informal human reasoning, cognitive processes, and social interactions. The concept of proof-event was coined by Goguen who described mathematical proof as a social event that takes place in space and time, designed to cover not only “traditional” formal proofs but all kinds of proofs including incomplete or purported proofs. In real-life cognitive processes, informal human reasoning and social aspects play a significant role. Our approach attempts to make proof-events more comprehensive to express the complete trajectory of a mathematical proofevent, including formal and informal proving steps, until the ultimate validation of the proving outcome. Thus, we present an extended version of proof-event calculus named Argumentation-based Proof-Event Calculus (APEC) which is built on the argumentation theories of Pollock, Toulmin, and Kakas designed to capture the internal and external structure of a collaborative mathematical practice. The second part of this thesis demonstrates two scopes of implementation to highlight the applicability and the expressivity of logic-based approaches in real-life proving scenarios. The first scope concerns explicit mathematical proving practices, which can be applied for an in-depth analysis of the internal steps in a mathematical proof, as indicated in the paradigm of Zero Knowledge Proofs, or for modeling a more external perspective to highlight the social interactions and the progress during a multi-agent proving process, as illustrated in the cases of Mini-Polymath 4 project and Fermat’s Last Theorem. The second scope is for implicit proving processes encoded in legal and ethical aspects of medical devices and wearable robots. Legal-AI models in the medical sector are presented through logic-based systems, where a legal text is represented by rules that can express legal arguments and exceptions and can provide explanations as audit trails of how a particular conclusion was proved. The logic-based legal systems presented in this work are: the WeaRED, an ethical decision-making system on Wearable Robots’ data privacy; the AMeDC and the Medical Devices Rules systems regarding Medical Devices Regulation; and the ExosCE Rules system regarding the regulation of exoskeletons.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2022

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Almpani, Sofia
  • Αλμπάνη, Σοφία

Subjects

dc:subject × 22

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/51572

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Almpani, Sofia; Αλμπάνη, Σοφία. Argumentation and rule-based logic in mathematical proving and legal artificial intelligence applications. National Technical University of Athens (NTUA), 2022. http://hdl.handle.net/10442/hedi/51572