Fuzzy Labeling for Abstract Argumentation: An Empirical Evaluation - Archive ouverte HAL Access content directly
Conference Papers Year : 2016

Fuzzy Labeling for Abstract Argumentation: An Empirical Evaluation

Abstract

Argumentation frameworks have to be evaluated with respect to argumentation semantics to compute the set(s) of accepted arguments. In a previous approach, we proposed a fuzzy labeling algorithm for computing the (fuzzy) set of acceptable arguments, when the sources of the arguments in the argumentation framework are only partially trusted. The convergence of the algorithm was proved, and the convergence speed was estimated to be linear, as it is generally the case with iterative methods. In this paper, we provide an experimental validation of this algorithm with the aim of carrying out an empirical evaluation of its performance on a benchmark of argumentation graphs. Results show the satisfactory performance of our algorithm, even on complex graph structures as those present in our benchmark.
Fichier principal
Vignette du fichier
SUM2016.pdf (3.64 Mo) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01377557 , version 1 (07-10-2016)

Identifiers

Cite

Célia da Costa Pereira, Mauro Dragoni, Andrea G. B. Tettamanzi, Serena Villata. Fuzzy Labeling for Abstract Argumentation: An Empirical Evaluation. Tenth International Conference on Scalable Uncertainty Management (SUM 2016) , Sep 2016, Nice, France. pp.126 - 139, ⟨10.1007/978-3-319-45856-4_9⟩. ⟨hal-01377557⟩
195 View
128 Download

Altmetric

Share

Gmail Facebook X LinkedIn More