The trend of individualization calls for personalized and accessible public transport services. Existing mobility services often fail to meet these requirements, as personal and context-sensitive travel needs are hardly considered. To enable services to better fit the needs of individuals, we developed a user segmentation method linking traveler characteristics (e.g., age, gender, public transport subscription) with route selection preferences. The segmentation method was comprised of a three-stage modular process. In the first module, user travel data is preprocessed and transformed for cluster analysis. This step addresses the challenges of high dimensionality and mixed data types by applying techniques such as principal component analysis, factor analysis for mixed data or simple feature selection. The choice of technique depended on the extent of the challenges. In the second module, unsupervised machine learning techniques were applied to cluster users based on the extracted components. The exact choice of the clustering technique depends on the data characteristics generated by the previous step. In the third module, relationships between user groups and route choices were analyzed. Routes were systematically classified based on attributes such as fastest connection or fewest transfers. Different statistical analyses were then conducted to examine the relationships between user clusters and the classified routes. The connections between the users clustered in step two and the classified routes were then analyzed using regression analyses. Applying the method on a dataset revealed that there is a difference in route selection between the clusters.