Clustering Analysis of Tourism and Crime in Barcelona from 2011 to 2022

Abstract

Barcelona, Spain, is one of the most popular cities for tourists in the world, which is visited not only during high season by local and by international tourists. Similarly to other cities that receive many tourists, residents of Barcelona usually perceive an impact on regular criminality levels during the year depending on tourism. This paper presents a clustering analysis approach to identify trends on the relationship between crime and tourism in Barcelona between 2011 and 2022. The clustering analysis presented considers three clustering methods: (i) K-Means, (ii) Hierarchical clustering, and (iii) DBSCAN. The rationale for using these techniques was to compare the overlaps and variations in the results obtained. Results were analysed in three different ways: (i) showing clusters as a layer over crime and tourism variables visually in maps, (ii) using cluster boxplots, and (iii) comparing mean values of individual cluster types for crime and tourism variables. The results obtained show that clusters tend to form around areas in which the volume of crime and tourism is either at its highest or lowest. Future work will consider other types of analysis such as regression, correlation and data visualizations.