Time-Dependent Tourist Tour Planning with Adjustable Profits.

Planning a tourist trip in a foreign city can be a complex undertaking: when selecting the attractions and choosing visit order and visit durations, opening hours as well as the public transit timetable need to be considered. Additionally, when planning trips for multiple days, it is desirable to avoid redundancy. Since the attractiveness of activities such as shopping or sightseeing depends on personal preferences, there is no one-size-fits-all solution to this problem. We propose several realistic extensions to the Time-Dependent Team Orienteering Problem with Time Windows (TDTOPTW) which are relevant in practice and present the first MILP representation of it. Furthermore, we propose a problem-specific preprocessing step which enables fast heuristic (iterated local search) and exact (mixed-integer linear programming) personalized trip-planning for tourists. Experimental results for the city of Berlin show that the approach is feasible in practice.