The definition of an index to synthesize the tourism appeal of a holiday destination is complex due to the effect of different aspects, such as the economic, socio-demographic, cultural, geographical ones regarding both demand and supply-side. In this paper, several spatially-referenced factors, related to the tourism attractiveness, are analyzed through the geographically weighted principal components analysis (GWPCA). However, the automatic setting of its kernel bandwidth, often used in practice, provides sometimes a not satisfactory result, since it is too small, with possible abrupt variation in the spatial domain, or too large, disregarding the spatial dependence. For this reason, a new approach, based on variography and minimum spatial correlation distance characterizing the covariates, is proposed for the GWPCA. A comparison with respect to the outputs of the GWPCA, based on automatic fitting, is discussed. Moreover, tourism composite spatial indicators are developed in order to support the policy makers in planning possible actions to boost tourism over the region of interest.

Nowadays, Artificial Intelligence (AI) is playing a rapidly increasing role in several fields of research and in almost all sectors of real life. However, few studies have assessed the effects of AI applications on training needs. This paper proposes an innovative multilevel modeling in order to investigate Awareness, Attitude and Trust towards AI and their reflections on learning needs. In particular, it is shown how a machine learning variable selection algorithm can support the definition of the optimal subset of all relevant covariates with respect to the outcome variable and improve the multilevel model performance for estimating the probability of educational needs. Thus, starting from a complex web survey to European citizens distributed in eight countries, the estimation of a multilevel binary model, defined on the basis of covariates selected through the Boruta random forest algorithm, is proposed. A discussion on the gender differences of the related estimated multilevel logit models is presented. A sensitivity analysis is also included in order to assess the prediction accuracy of the proposed multilevel logit modeling.