Towards Context-Aware Recommender Systems for Tourists

Abstract

The data warehouse investments of online retail and media providers have made them leaders in exploiting pattern-finding techniques which are revolutionising financial services, health care, tourism and leisure. Machine Learning plays a pivotal role in this revolution, enabling the analysis of massive datasets to uncover complex trends and generate personalised insights with the widespread use of Recommender Systems (RS). To ensure the continued effectiveness and innovation in RS algorithms, it is crucial to actively seek out diverse datasets for developing and evaluating personalisation solutions. This thesis contributes to the RS field by investigating novel applications of RS, emphasising trust-based mechanisms for Group Recommender Systems (GRS), contextual factors within Next-POI (Point of Interest) algorithms, and route recommendation techniques within the dynamic tourism domain.

Specifically, this research addresses limitations identified in traditional RS systems through a unique multi-faceted approach. Firstly, a YouTube API-based Group Decision-Making (GDM) model is explored, where we explore the effect of a trust-driven aggregation on interpersonal dynamics for optimising recommendations in tourism planning. Secondly, a Context-Aware Recommender System (CARS) for restaurant deal recommendations is presented, balancing factors such as user preferences, popularity, and location. This work directly showcases the improvement when tested on a dataset with over 300K transactions. Finally, focusing on personalised Next-POI and route generation in tourism, a CARS algorithm is developed that actively incorporates real-time contextual data such as location, weather, and business hours for creating adaptive solutions.

Overall, this thesis demonstrates the innovative potential of RSs and CARSs within the tourism sector, with actual impacts on tourists’ decision-making processes and their overall travel experiences.

Date of Award	16 May 2024
Original language	English
Awarding Institution	University of Bristol
Sponsors	Turkish Ministry of National Education
Supervisor	Jonathan Lawry (Supervisor) & Seth Bullock (Supervisor)