The bi-objective orienteering problem with hotel selection: an integrated text mining optimisation approach

To assist the tourist in planning a more desirable trip according to his/her preferences, a variety of elements including accommodation, sightseeing locations, cost budget, and time constraints should be considered. To provide such a through planning, the Orienteering Problem with Hotel Selection (OPHS) has been recently introduced in the literature. However the current OPHS variants do not use hotels’ aspects in their modelling thereby are not able to provide a comprehensive planning. In this work, we address this gap by introducing a bi-objective OPHS which maximises the utility of the sightseeing, as well as the total weighted scores of hotel selection and provides multi day tourist trip plans. Moreover, as cities become smarter, colossal data can be used to facilitate tourist trip planning according to the preferences of tourists. In this research, we leverage the power of relevant external data thereby providing effective tourist plans. In particular, a decision-making framework is proposed for the multi-day tourist planning with hotel selection. This Integrated Text-mining Optimisation Framework consists of three modules. First, the accommodations’ scoring is measured by Aspect-Based Sentiment Analysis on existing reviews, looking at fifteen quality aspects of hotels. Second, four different tourist segments are considered, and the priority for each of the hotel’s aspects in each segment is calculated using the Best-Worst Method. The third module is solving the bi-objective OPHS to provide the desired plans for the tourist. To evaluate the applicability of the method, a real-world data set is analysed.