Big Data Analytics and Predictive Modelling of Tourist Activity in Ukraine

Abstract

Design/ Methodology/ Approach: The methodological basis of the work is a combination of quantitative and qualitative methods. The study is based on an analysis of official statistical data, digital platforms, social networks, and user search activity for 2018–2025. The sample included more than 120,000 digital records, geotags, and tourist observations, aggregated by monthly time series. Regression modelling, machine learning, spatial geo -analysis, as well as analysis of tourist reviews in social networks. Research Limitation: The study's tourist flow estimates are therefore partially contingent upon the quality and ethical standing of data sources whose limitations may not be entirely visible within the scope of the research, and this represents a structural constraint on the epistemic confidence with which the quantitative findings of this study may be advanced and applied within both academic and policy-facing contexts. Findings: Based on data processing for 2018-2025, key factors influencing the formation of tourist flows were identified: seasonality, users' information activity on the network, transport mobility, socio-economic indicators, and natural and climatic conditions. The built-in forecasting model achieved an average prediction accuracy of 8.4%, exceeding the results of traditional statistical methods by 27%. Practical Implication: The research demonstrates the feasibility of implementing Big Data technologies. Data analytics for the tourism management system of Ukraine, in particular, the creation of a National Tourism Analytics System for monitoring, planning, and developing regional destinations. Social Implication: The use of Big Data in forecasting tourist flows improves the efficiency of tourism infrastructure management, supports the development of regional tourism, and ensures a more balanced distribution of tourist flows among destinations. This is especially important for Ukraine in the context of the tourism industry's post-crisis and post-war recovery. Originality / Value: The scientific novelty of the study lies in integrating machine learning methods, social network analysis, Google Trends search activity, and geoanalytics into a single model to forecast tourist flows in Ukraine amid digital transformation and post-war recovery.