Modeling evolutionary responses in crowdsourcing MCQ using belief function theory - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Modeling evolutionary responses in crowdsourcing MCQ using belief function theory

Résumé

Crowdsourcing is a widespread method for outsourcing tasks to a crowd of contributors. These simple tasks are often formulated by multiple choice questionaries (MCQs) to which the contributor has to give a precise answer. We hypothesize that offering the contributor to fill in an imperfect answer and to evolve it is more profitable than a single precise answer. In this paper, we propose a model for the evolutionary answers of contributors to MCQs in crowdsourcing platforms. In order to realize this modeling we use the theory of belief functions. The model and experiments conducted on real data from crowdsourcing campaigns are presented in this paper. Our experiments show that modelling evolutionary responses using consonant mass functions improves the quality of the results obtained when aggregating responses compared with majority voting.
Fichier principal
Vignette du fichier
KES_THIERRY-2.pdf (2.69 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04214522 , version 1 (06-02-2024)

Licence

Paternité

Identifiants

  • HAL Id : hal-04214522 , version 1

Citer

Constance Thierry, Arnaud Martin, Yolande Le Gall, Jean-Christophe Dubois. Modeling evolutionary responses in crowdsourcing MCQ using belief function theory. 27th International Conference on Knowledge-Based and Intelligent Information & Engineering Systems (KES 2023), Sep 2023, Athens, Greece. ⟨hal-04214522⟩
12 Consultations
3 Téléchargements

Partager

Gmail Facebook X LinkedIn More