Research Seminar by TRAN Nhu Thuat and DONG Viet Hoang

Talk #1: VARIUM: Variational Autoencoder for Multi-Interest Representation with Inter-User Memory
by TRAN Nhu Thuat, PhD Candidate

Variational AutoEncoder (VAE)-based frameworks effectively model multiple user interest factors but struggle to facilitate interest sharing across users. To address this, we introduce an inter-user memory mechanism that unsupervisedly discovers latent interest sharing within a VAE framework. Our memory consists of prototypes, each representing a shared user interest, which are jointly trained with the backbone model. User-specific interest factors first capture intra-user preferences, then query the memory to retrieve inter-user interest clues through a sequential attention-transformation process. These retrieved clues refine interest representations, enhancing recommendation performance. Extensive experiments on real-world datasets demonstrate the effectiveness of our approach, with qualitative analysis providing further insights into its inner workings.

Talk #2: A Contrastive Framework with User, Item and Review Alignment for Recommendation
by DONG Viet Hoang, PhD Candidate

Learning effective latent representations for users and items is the cornerstone of recommender systems. Traditional approaches rely on user-item interaction data to map users and items into a shared latent space, but the sparsity of interactions often poses challenges. While leveraging user reviews could mitigate this sparsity, existing review-aware recommendation models often exhibit two key limitations. First, they typically rely on reviews as additional features, but reviews are not universal, with many users and items lacking them. Second, such approaches do not integrate reviews into the user-item space, leading to potential divergence or inconsistency among user, item, and review representations. To overcome these limitations, our work introduces a Review-centric Contrastive Alignment Framework for Recommendation (ReCAFR), which incorporates reviews into the core learning process, ensuring alignment among user, item, and review representations within a unified space. Specifically, we leverage two self-supervised contrastive strategies that not only exploit review-based augmentation to alleviate sparsity, but also align the tripartite representations to enhance robustness. Empirical studies on public benchmark datasets demonstrate the effectiveness and robustness of ReCAFR.

TRAN Nhu Thuat is a PhD Candidate in Computer Science at the SMU School of Computing and Information Systems, supervised by Prof. Hady W. Lauw. His research focuses on disentangling user preferences towards interpretable recommender systems.

DONG Viet Hoang is a PhD Candidate in Computer Science at the SMU School of Computing and Information Systems, supervised by Assistant Prof. FANG Yuan and Prof. Hady W. Lauw. His research focuses on enhance user and item preferences with side information for recommender systems.