Comparison of hard and probabilistic evidence in bayesian model

Abstract

Bayesian networks are powerful tools for probabilistic reasoning with uncertain evidences. Evidence originates from information based on the variables of observation. In this paper, we focus on two types of evidences: hard evidence and probabilistic evidence. We were interested in updating an evidence represented by a Bayesian model. This paper presents the application of probabilistic evidence in an adaptive user interface. Then, we compare the Bayesian model using probabilistic evidence with the Bayesian model using hard evidence.