Reasoning over Assumption-Based Argumentation Frameworks via Direct Answer Set Programming Encodings

Authors: Tuomo Lehtonen, Johannes P. Wallner, Matti Järvisalo2938-2945

AAAI 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We present empirical results comparing the performance of the state-of-the-art ASP solver Clingo on our ASP encodings to currently available systems for ABA and ABA+.
Researcher Affiliation Academia Tuomo Lehtonen University of Helsinki, Finland Johannes P. Wallner TU Wien, Austria Matti J arvisalo University of Helsinki, Finland
Pseudocode Yes The paper includes 'Listing 1: Module πcommon', 'Listing 2: Module πadm', and 'Listing 3: Module πgrd' which provide structured Answer Set Programming (ASP) code that describes the algorithms used.
Open Source Code No The paper mentions systems like abagraph and aba2af, and tools like Clingo and SICStus Prolog, which are used or compared against, but does not provide a link or explicit statement that the authors' own ASP encodings or implementation code for this paper are open-source or publicly available.
Open Datasets Yes We used the ABA frameworks (which contain up to 90 sentences) and queries used by Craven and Toni (2016) and Lehtonen, Wallner, and J arvisalo (2017) in experiments on abagraph and aba2af ( http://robertcraven.org/proarg/ experiments.html)
Dataset Splits No The paper describes how instances were generated ('generated three frameworks for each number of sentences' and 'generated preferences by choosing a random permutation... with two fixed probabilities'), but it does not specify explicit train/validation/test dataset splits with percentages or sample counts.
Hardware Specification Yes The experiments were run on 2.83-GHz Intel Xeon E5440 quad-core machines with 32-GB RAM under Linux using a 600-second time limit per instance.
Software Dependencies Yes We used Clingo v5.2.2 (Gebser et al. 2016) as the ASP solver, and SICStus Prolog v4.4.1 for abagraph.
Experiment Setup Yes The experiments were run on ... using a 600-second time limit per instance. ... We generated preferences by choosing a random permutation (ai)0<i n of the assumptions, and for each j < i, set ai to be preferred to aj with two fixed probabilities, 15% and 40%.