E-commerce sellers aim to better understand trust formation mechanisms and assess the level of initial trust created by their sites on potential clients. In this work, a survey is conducted to assess websites characteristics as viewed by the respondents and reveal their perceptions. Based on the survey data, we resort to rules classifiers to provide new insights on website trust and suggest specific recommendations for e-commerce vendors. Decision trees and rough sets are used to this end. In addition, we propose a heuristic aiming to derive simpler classifiers, taking into account their predictive ability and the parsimony of rules’ sets.

--

trust,e-commerce,rough sets,decision trees