The use of knowledge graph for recommendation explanation

• Main Author: Yap, Desmond Qing Yang
• Other Authors: Fan Xiuyi
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
• Online Access: https://hdl.handle.net/10356/166703
• Institution: Nanyang Technological University

This research paper proposes an approach for explainable recommendation systems that incorporates both knowledge graphs and curiosity-driven exploration. It utilises a knowledge graph to provide transparent and intuitive explanations for recommendations, while also incorporating curiosity-driven exploration to encourage users to discover novel and interesting items. The system employs a knowledge graph that stores structured information about items and user pairs to provide explanations for the recommendations. The knowledge graph is integrated with a recommendation algorithm that uses a hybrid approach of content-based filtering and collaborative filtering to make recommendations while also incorporating a curiosity-driven exploration by suggesting items that are less popular but might be of interest to the user. In conclusion, this research paper demonstrates the potential of integrating recommendation systems with knowledge graphs and curiosity-driven exploration for explainable recommendations. The proposed approach provides a transparent and comprehensive way for users to discover new and interesting items, which is of great importance for many applications.