In thesis an approach for analyzing of visiting tourist destinations with graph algorithms on publicly available data was built. The goal of analysis is to find groups of destinations frequently visited together that can be a base for a joint marketing by destinations. Publicly available data can be tourists' posts related to connected to visit of tourist destination from different social networks or other sources. The implemented solution uses data gathered from public tourists' posts on TripAdvisor as input to generate co-occurrence graph of destinations. Weight of edge represents sum of trips, where both destinations were visited. For analysis of destinations frequently visited together we build on the methods, that are used for market basket analysis (MBA). We used Apriori algorithm and two algorithms for community detection: Louvain and Infomap. The most useful results were obtained by using Infomap. Data filtering enables us to observe different segments of guests based on their profile. The largest difference between communities was detected when using segmentation based on guests' country of origin.