A5039641800- Recommender Systems and Techniques
- Topic Modeling
- Advanced Graph Neural Networks
- Image Retrieval and Classification Techniques
- Music and Audio Processing
- Natural Language Processing Techniques
- Text and Document Classification Technologies
- Semantic Web and Ontologies
University of Hong Kong
2023-2024
Graph neural networks (GNNs) have recently emerged as an effective collaborative filtering (CF) approaches for recommender systems. The key idea of GNN-based systems is to recursively perform message passing along user-item interaction edges refine encoded embeddings, relying on sufficient and high-quality training data. However, user behavior data in practical recommendation scenarios often noisy exhibits skewed distribution. To address these issues, some approaches, such SGL, leverage...
Knowledge Graphs (KGs) have emerged as invaluable resources for enriching recommendation systems by providing a wealth of factual information and capturing semantic relationships among items. Leveraging KGs can significantly enhance performance. However, not all relations within KG are equally relevant or beneficial the target task. In fact, certain item-entity connections may introduce noise lack informative value, thus potentially misleading our understanding user preferences. To bridge...
Multimedia online platforms (e.g., Amazon, TikTok) have greatly benefited from the incorporation of multimedia visual, textual, and acoustic) content into their personal recommender systems. These modalities provide intuitive semantics that facilitate modality-aware user preference modeling. However, two key challenges in multi-modal recommenders remain unresolved: i) The introduction encoders with a large number additional parameters causes overfitting, given high-dimensional features...
The rise of online multi-modal sharing platforms like TikTok and YouTube has enabled personalized recommender systems to incorporate multiple modalities (such as visual, textual, acoustic) into user representations. However, addressing the challenge data sparsity in these remains a key issue. To address this limitation, recent research introduced self-supervised learning techniques enhance systems. methods often rely on simplistic random augmentation or intuitive cross-view information,...
Modern recommender systems aim to deeply understand users' complex preferences through their past interactions. While deep collaborative filtering approaches using Graph Neural Networks (GNNs) excel at capturing user-item relationships, effectiveness is limited when handling sparse data or zero-shot scenarios, primarily due constraints in ID-based embedding functions. To address these challenges, we propose a model-agnostic recommendation instruction-tuning paradigm that seamlessly...
Knowledge Graphs (KGs) have emerged as invaluable resources for enriching recommendation systems by providing a wealth of factual information and capturing semantic relationships among items. Leveraging KGs can significantly enhance performance. However, not all relations within KG are equally relevant or beneficial the target task. In fact, certain item-entity connections may introduce noise lack informative value, thus potentially misleading our understanding user preferences. To bridge...
Graph neural networks (GNNs) have recently emerged as an effective collaborative filtering (CF) approaches for recommender systems. The key idea of GNN-based systems is to recursively perform message passing along user-item interaction edges refine encoded embeddings, relying on sufficient and high-quality training data. However, user behavior data in practical recommendation scenarios often noisy exhibits skewed distribution. To address these issues, some approaches, such SGL, leverage...