Building confidence measures for tourist destination choice

Abstract Reviews, social media, web sites and recommendations from trusted inner circle opinions are essential to tourists in the travel planning process, but the large amount of information they now make available can increase the difficulty of choosing a destination. Determining how improvements in data quality and in the variety of tourist information sources will affect the probability of correct situational detection, and help in optimizing the overall decision, is vital. This paper presents a generalized method for estimating levels of individual source reliability and assessing the confidence level in a combined output. Initial reliability levels of tourist information sources are augmented by: (i) combining measures of multiple sources, (ii) incorporating truth reinforcement for related elements and (iii) incorporating the importance of individual elements for determining the truth probability for the whole. The result is a measure of confidence in system output based on establishing links among the truth values of inputs.