Examining how predictive analytics may be used in smart tourism frameworks to anticipate and improve sports tourism visitor experiences is the aim of the research. The paper aims to examine four predictive models – logistic regression, support vector machine (SVM), artificial neural network (ANN) and random forest – to forecast visitors' intent to propose sports venues using the Cross-Industry Standard Process for Data Mining (CRISP-DM) technique. The research highlights logistic regression as the best model, finds important demographic and contextual variables that impact proposals and provides helpful recommendations for improving consumer satisfaction and strategic planning in sports tourism.
The CRISP-DM approach is used in this research to operationalize predictive analytics in sports tourism. Four classification models – logistic regression, SVM, ANN and random forest – were used to examine a dataset that has visitor records. Demographic, behavioral, contextual and experience-based features were among the variables. Metrics like area under the curve (AUC), accuracy, precision, recall, F1 score and Matthews correlation coefficient (MCC) were used to assess the preprocessed data, which had been divided into 60% training and 40% testing sets. In a smart tourism framework, the most effective methodology was logistic regression, which correctly predicted travelers' intention to suggest sporting venues.
AUC, accuracy, precision, recall, F1 score and MCC = 1.000 were all ideal scores for logistic regression, which the research showed to perform more accurately than other models. Tourists' intention to suggest was influenced by a number of important features, including gender, country, event type, ticket type and season of travel. The venue was less likely to be recommended by male tourists, general ticket holders and visitors of summer events. Conversely, those who attended tennis matches, had very-important-person tickets and traveled during the winter shown more desire to suggest. These findings demonstrate how predictive analytics may improve decision-making and customize sports tourist experiences.
In a field with little empirical research, sports tourism, this study combines predictive analytics in an unconventional manner inside a smart tourism framework. The research provides a data-driven strategy for comprehending the recommendation behavior of tourists by using the CRISP-DM methodology and numerous machine learning models. For tourist planners and event organizers, the use of comparative model analysis and actual visitor data offers beneficial findings. The results not only pinpoint important factors that influence satisfaction but also direct focused tactics to enhance visitor experience and service delivery. By showing how predictive modeling may improve strategic choices in the ever-changing sports tourism industry, this research closes a significant gap.
