Knowledge graph construction and recommender system development of tourism in Singapore

Abstract

This research and development project presents a comprehensive investigation into the development and implementation of a knowledge-graph-based recommender system tailored for urban tourism. The recommender system is powered by a recommender engine developed in this project, which utilizes a combination of data mining algorithms, such as heuristic methods, content-based filtering, collaborative filtering, and ensemble learning techniques to generate personalized recommendations for tourist points of interest (POI) in Singapore. A key focus of the study is the evaluation of individual data mining algorithms and ensemble learning strategies to provide insights into their performance across various metrics such as precision, recall, and coverage score. The research identifies the strengths and limitations of each approach, highlighting the importance of a user-centric design and the challenges posed by data and resource constraints. Future work is outlined, including advancements in ensemble learning, database scaling, and user feedback analysis. Overall, the project contributes to the field of knowledge graph and recommender systems by offering a practical framework for developing knowledge-graph-based recommender systems application in urban tourism.