Rating credibility of sources for profiling risk and business context of service requests - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2014

Rating credibility of sources for profiling risk and business context of service requests

Résumé

Attributes credibility, trustiness and accuracy are often considered, but not incorporated in decision making algorithms. In this paper a credibility-based risk/business context model is proposed, to profile communication service requests using sources' credibility, as well as observed intensity and customizable policy-based prioritization. The paper proves that this Credibility-based approach can cope with the complexity and uncertainty of context models, which Bayes probabilities cannot, and with degrees of discordance and uncertainty, which DST cannot. The paper provides reviews of concepts and techniques to determine credibility from its components and aggregate corroborated credibility within observations and ultimately key factors of the model. The paper tests and proves the effectiveness of the modelled credibility, intensity and policy, and shows that incorporating Credibility is a better way to make informed decisions
Fichier non déposé

Dates et versions

hal-01334171 , version 1 (20-06-2016)

Identifiants

Citer

Rebecca Copeland, Noel Crespi. Rating credibility of sources for profiling risk and business context of service requests. SCC 2014 : IEEE International Conference on Services Computing, Jun 2014, Anchorage, United States. pp.416 - 423, ⟨10.1109/SCC.2014.62⟩. ⟨hal-01334171⟩
211 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More