Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

In this thesis, the potential of call detail record data (CDR data) is examined to study touristic long-distance movement in a spatial context. The extraction of domestic tourism movements based on a CDR dataset of Orange Labs is explored, containing logs of 5 months for 65.000 users in the 31 largest cities of France. A step-by-step plan was designed to derive touristic destinations from this CDR dataset. First, based on a methodology developed by Janzen et al. (2016), long-distance movements were extracted from the CDR data. Secondly, a proxy was implemented based on the duration and starting date of the long-distance movements to obtain a set of “touristic” long-distance trips. Third, an algorithm, namely the DCR algorithm, was developed to indicate the location of the destination from a touristic long-distance trip. An interactive validation was created to determine the optimal parameters and the sensitivity of the algorithm towards parameter choices. This step-by-step plan resulted in the detection of more than 18.000 destinations and a wide scale estimation of touristic origin-destination fluxes within France. Next, socio-economic and environmentally related variables were used as input for two types of logistic multivariate regression models, namely (1) touristic pull and (2) touristic push models, which had as aim to predict the obtained spatial pattern of domestic tourism in France. Although the touristic push models were not significant and the predictive values of the pull models were rather moderate, it was possible to find significant pull factors and to derive general tourism patterns from the pull models. The resulting tourism patterns are corresponding to the results of classical data sources like surveys and questionnaires. Furthermore, the patterns derived from CDR data can even add extra value, because tourism fluxes are extracted on a more detailed scale and in a more cost efficient way. Moreover, the fluxes derived from CDR data are more representative, because the size of the used CDR dataset is larger than the average sample size of surveys and questionnaires. Therefore, this work can be of importance for the tourism industry and regional policies.