Probabilistic Modelling of Influences on Travel Decision Making

For any tourism organisation or company, it is a necessity to know about the factors that are influencing tourists’ travel decisions. The question, however, is how to model and represent heterogeneous influence factors in a way that a) human decision makers can easily understand, and b) allows for computer-based simulation and diagnoses to aid decision making. Currently, tourism suffers from meaningful and practically applicable representations of complex relations among influence factors stemming from different domains. This paper investigates Bayesian networks for modelling mutual influences of factors originating from heterogeneous data sources including tourism experts, and the integration of associated uncertainties in a single model. The authors are elaborating several development alternatives for the creation of a Bayesian network-based tourism knowledge model. Using this model, tourism professionals will be able to perform interactive decision analyses for determining, e.g., how to spend marketing budget most efficiently.