Towards a numerical, agent-based, behaviour analysis: The case of tourism

Abstract

We discuss who should be in charge of providing data relevant to marketing segmentation for the tourism industry. We review the most commonly found consumer behavioural models and discuss the difficulties of their integration within an information system. They will be oppose to a novel approach in marketing segmentation, based on outgoings analysis. We use agent-modelling techniques, based on cellular automaton rules and stochastic processes to implement our model and generate sales data. We then present our algorithm to identify similarly behaved tourists, showing that the commonly used "nationality" variable for segments discrimination is not efficient. We conclude with some test runs results discussion and possible further research tracks.