A Trust Model for Ubiquitous Healthcare Environment on the Basis of Adaptable Fuzzy-Probabilistic Inference System

Trust is considered to be a determinant on psychologist selection which can ensure patient satisfaction. Hence, trust concept is essential to be introduced into ubiquitous healthcare (UH) environment oriented on patients with anxiety disorders. This is accomplished by trust model estimating psychologists' trustworthiness, a priory to service delivery, with the use of patient's and his/her acquaintances testimonies, i.e., personal interaction experience and reputation (R). In this paper, a trust model is proposed to be materialized via an adaptable cloud inference system (ACIS) that performs trust value (TV) estimation. Taking advantage of a cloud theory, the introduced ACIS estimates TVs via fuzzy-probabilistic reasoning incorporating a cloud relation operator (soft AND) which is proposed to be tuned by trust information sources consistency and coherency. Theoretical analysis along with comparative study conducted within MATLAB environment and experimental investigation verify the effectiveness of the proposed ACIS materialization under different conditions. Especially, the innovative features of ACIS enable TV to be estimated with 45.5% and 62% on average higher accuracy to that providing state-of-the-art trust models, within clean environment and under the influence of large-scale collusive malicious attacks, respectively. The enhanced robustness permits the untrustworthy UH providers to be discriminated with true positive rate at the range of 0.9 although 40% of R testimonies are erroneous. Finally, experimental investigation validates that the adoption of the proposed trust model for psychologists trustworthiness estimation facilitates patient satisfaction to be achieved into UH environment.