Mimic Visiting Styles by Using a Statistical Approach in a Cultural Event Case Study

Classifying the behaviours of users in the cultural spaces with the aim to infer knowledge about the event fruition is a fascinating challenge. In this paper, starting from real data, we are interested in predicting the user dynamics related to the interaction of a spectator with artworks and with the available technologies. Clustering techniques are preliminary used to find groups that reflect visiting styles. Accordingly with this, we assume that visitors are previously classified. The start-up dynamical process underlying classification can be affected by several errors. Here we adopt a powerful statistical method to predict the visiting style dynamics of spectators. Finally, numerical experiments confirm that it is possible to predict the visitors' behaviour with good results.