Multiple criteria fake reviews detection using belief function theory - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2020

Multiple criteria fake reviews detection using belief function theory

Résumé

Checking online reviews before making a purchase becomes a permanent habit. Hence, online consumer reviews, product and services play an increasingly spreading role in consumer purchasing decisions. Unfortunately, the importance of advertising and the attraction of profit have led to the appearance of fake reviews in order to mislead readers. Considering that the reviews are generally imperfect, the spam reviews detection becomes one of the most important problems. To tackle this problem, we propose a new method of multi-criteria fake reviews under belief function theory. This approach treats the uncertainty in the rating reviewers' given to multiple evaluation criteria, takes into account the similarity between all provided reviews and deals with missing data. We evaluate our method through artificial datasets. Then, we use a real dataset to validate it. The results prove that the proposed approach is a useful solution for the fake reviews detection problem.
Fichier principal
Vignette du fichier
MCBFRD-ISDA2018.pdf (313.33 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03643831 , version 1 (16-04-2022)

Identifiants

Citer

Malika Ben Khalifa, Zied Elouedi, Eric Lefevre. Multiple criteria fake reviews detection using belief function theory. International Conference on Intelligent Systems Design and Applications, ISDA'2018, Dec 2018, Vellore, India. pp.315-324, ⟨10.1007/978-3-030-16657-1_29⟩. ⟨hal-03643831⟩

Collections

UNIV-ARTOIS LGI2A
33 Consultations
38 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More