MSc Project: On Tour

In this MSc thesis project, I introduce a variety of data-driven models for recommending both city destinations and within-city points of interest to tourists. The models are implemented with a novel dataset of travel histories, derived from social media data, which is larger by size and scope than in prior work. All proposed models outperform simple baselines in cross-validation experiments, with the strongest variants reliably including tourists’ true movements among their top recommendations.

See here for the GitHub repository.