Leveraging Heterogeneous GNNs for User Behaviour Analysis and Node Classification in Movie Recommender Systems

Nayak, Nishant (2025) Leveraging Heterogeneous GNNs for User Behaviour Analysis and Node Classification in Movie Recommender Systems. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (592kB) | Preview

Abstract

In this study, we examine enhancements to movie recommendation systems utilizing hybrid heterogeneous graph neural networks (GNNs) to effectively model complex user-movie dynamics and genre associations. A graph-centric framework is introduced, utilizing a real-world dataset to derive embeddings for users and movies, resulting in improved recommendation accuracy. The primary emphasis is on differentiating between “normal” users, who exhibit moderate movie ratings, and the outlier “top” users with high activity levels. Through comprehensive histogram evaluations and embedding visualizations, we illustrate that the majority of users are classified as “normal,” whose preference patterns should be prioritized for model optimization. The model's interpretability is elucidated through t-SNE, cluster analysis, and genre heatmaps. Our results indicate that hybrid GNN architectures possess strong generalization capabilities, effectively accommodating diverse taste profiles and delivering recommendations for both mainstream and extreme user behavior. This research promotes equitable and pragmatic personalization within movie recommendation systems.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Horn, Christian UNSPECIFIED
Uncontrolled Keywords:	Movie Recommendation Systems; Recommender Systems; Graph Neural Networks (GNNs); Heterogeneous Graphs; Hybrid GNN Architectures; User Embeddings; Movie Embeddings; User Clustering; User Behaviour Analysis; Outlier Detection; User Segmentation; t-SNE Visualization; Genre Preferences; Cluster Analysis; Collaborative Filtering (CF); Matrix Factorization (MF)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science H Social Sciences > HD Industries. Land use. Labor > Specific Industries > Film Industry
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	02 Jul 2026 14:07
Last Modified:	02 Jul 2026 14:07
URI:	https://norma.ncirl.ie/id/eprint/9441

Actions (login required)

View Item