The tourist trip design problem with POI categories via an Expectation-Maximization based method
Introduction
Figure 1: The schema of the proposed system architecture.
We propose awe propose an efficient deterministic method based on Expectation-Maximization (EM) to solve the challenging problem of the tourist trip design or Personalized Itinerary Recommendation (PIR) with POI categories. PIR aims to recommend a personalized tour that consists of a sequence of Points of Interest (POIs), which maximizes user satisfaction and adheres to user time budget constraints (see Figs. 1,2).
Methodology
This framework mainly focuses on the POIs sequence selection problem exploiting the personalized POI recommendations provided by a recommender system.
The proposed method sequentially solves the PIR problem by providing in each step the POI that is expected to maximize a suitable objective function, taking into account user satisfaction, user time budget, POIs opening hours, POIs category constraints and spatial constraints (e.g. start and end point, POIs locations, etc).
Fig. 2. A map of 16 POIs, where each POI is drawn by a circle. The size and the color of a POI correspond to the duration of the visit and the gained user satisfaction, respectively. The category of each POI is shown in parenthesis. The itinerary is indicated with a red line. The timetable of the personalized itinerary (bottom right).
Experiments - Downloads
You can download the matlab code of the PIREM, M-PIREM methods proposed in [1].
You can download the datasets of the method proposed in [1] from (LINK).
The code and extra material presented in [2] will be available after the publication of [2].
Related Publications
[1] C. Panagiotakis, E. Daskalaki H. Papadakis, and P. Fragopoulou, The tourist trip design problem with POI categories via an Expectation-Maximization based method, RecSys Workshop on Recommenders in Tourism, 2022.
[2] C. Panagiotakis, E. Daskalaki H. Papadakis, and P. Fragopoulou, An Expectation-Maximization framework for Personalized Itinerary Recommendation with POI Categories, Algorithms, 2022 (under review).