Javascript must be enabled for the correct page display

Implementing an Argumentation Game for ADFs under Preferred Semantics

Bouma, Marieke (2020) Implementing an Argumentation Game for ADFs under Preferred Semantics. Bachelor's Thesis, Artificial Intelligence.

[img]
Preview
Text
AI_BA_2020_Marieke_Bouma.pdf

Download (400kB) | Preview
[img] Text
toestemming.pdf
Restricted to Registered users only

Download (91kB)

Abstract

Formal argumentation can be used to model and evaluate reasoning, making it a powerful support for AI. Further, AI methods can help find answers to logical problems in formal argumentation. This project explores Abstract Dialectical Frameworks (ADFs), an argumentation formalism structured as a network of arguments. A propositional formula is attached to each argument, to indicate the conditions under which an argument can be accepted and to show the type of relation between arguments. One logical problem in ADFs is the credulous decision problem, which describes the question of whether there exists a set of truth values for the arguments under certain semantics in which a specific argument has a given value. Keshavarzi presents a discussion game algorithm between two players, which solves this problem under preferred semantics. In this game, the players take turns trying to find new information given a claim the previous player made about the truth values of arguments. Depth-First Search is applied to find the final set of truth values needed to solve the credulous decision problem. If DFS fails, then the initial claim is not satisfiable. This project consists of the implementation of this discussion game in Python.

Item Type: Thesis (Bachelor's Thesis)
Supervisor name: Verbrugge, L.C. and Keshavarzi Zafarghandi, A.
Degree programme: Artificial Intelligence
Thesis type: Bachelor's Thesis
Language: English
Date Deposited: 24 Jul 2020 14:31
Last Modified: 24 Jul 2020 14:31
URI: https://fse.studenttheses.ub.rug.nl/id/eprint/22860

Actions (login required)

View Item View Item