Session Based Recommendation

Session-based recommendation (SBR) systems aim to predict a user's next interaction based solely on their current browsing session, capturing short-term, dynamic preferences without relying on long-term user profiles. Recent research emphasizes improving SBR accuracy and diversity by incorporating inter-session relationships, leveraging social network information, and integrating diverse data modalities (e.g., text, images) using advanced architectures like graph neural networks (GNNs) and large language models (LLMs). These advancements enhance recommendation relevance and user experience, with implications for personalized online services across various domains, including e-commerce and news aggregation.