Learning Sparse Representations of Multimodal Content for Enhanced Cold Item Recommendation

• URL: http://arxiv.org/abs/2607.17184v1
• Date: Sun, 19 Jul 2026 10:45:57 GMT
• Title: Learning Sparse Representations of Multimodal Content for Enhanced Cold Item Recommendation
• Abstract summary: We argue that sparse embeddings have notable advantages over standard dense vectors in this content-based cold-start paradigm.<n>We show that the resulting sparse embeddings achieve significant improvements in cold-start recommendation accuracy over dense embeddings at considerably lower storage costs.
• Score: 5.958603849321135
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The scale and rapid growth of item catalogs in modern digital platforms present significant challenges to recommender system (RS) practitioners. Most RSs use embedding similarity to predict user-item preferences, but embedding storage and low-latency retrieval is challenging in industry-scale catalogs. Furthermore, newly added items do not have corresponding embeddings and cannot be recommended effectively; previous works often tackle this item cold-start problem by generating cold item representations from auxiliary content, such as images or descriptive text, so that user preferences can be predicted without historical interactions. In this paper, we argue that sparse embeddings have notable advantages over standard dense vectors in this content-based cold-start paradigm. We describe how existing cold-start training regimes can be adapted for sparse representation learning, and build on insights from linear attention to design a pre-sparsification activation technique that induces sharpness and denoising effects in learned item-item similarities. We show that the resulting sparse embeddings achieve significant improvements in cold-start recommendation accuracy over dense embeddings at considerably lower storage costs, especially for users with multiple interests. Through comprehensive experiments on four multimodal RS datasets, we also demonstrate the interpretability of sparse content embeddings and their robustness in the trade-off between size and accuracy.

Related papers

• Sparse Personalized Text Generation with Multi-Trajectory Reasoning [95.19317487799292]
  PAT (Personalization with Aligned Trajectories) is a reasoning framework for cold-start LLM personalization.<n>It first retrieves information along two complementary trajectories: writing-style cues from stylistically similar users and topic-specific context from preference-aligned users.<n>It then employs a reinforcement learning-based, iterative dual-reasoning mechanism that enables the LLM to jointly refine and integrate these signals.
  arXiv  Detail & Related papers  (2026-04-27T20:58:49Z)
• Sparse Contrastive Learning for Content-Based Cold Item Recommendation [5.958603849321135]
  Item cold-start is a pervasive challenge for collaborative filtering (CF) recommender systems.<n>Existing methods often train cold-start models by mapping auxiliary item content, such as images or text descriptions, into the embedding space of a CF model.<n>We propose to avoid this limitation with purely content-based modeling of cold items, i.e. without alignment with CF user or item embeddings.
  arXiv  Detail & Related papers  (2026-04-14T17:24:25Z)
• Efficient Cold-Start Recommendation via BPE Token-Level Embedding Initialization with LLM [1.1049570075807806]
  This paper presents an efficient cold-start recommendation strategy based on subword-level representations.<n>We obtain fine-grained token-level vectors that are aligned with the BPE vocabulary.<n>We show that using subword-aware embeddings yields better generalizability and is more interpretable.
  arXiv  Detail & Related papers  (2025-09-16T15:32:51Z)
• Let It Go? Not Quite: Addressing Item Cold Start in Sequential Recommendations with Content-Based Initialization [0.8437187555622164]
  We introduce a small trainable delta to frozen embeddings that enables the model to adapt item representations without letting them go too far from their original semantic structure.<n>This approach demonstrates consistent improvements across multiple datasets and modalities.
  arXiv  Detail & Related papers  (2025-07-25T17:57:31Z)
• Language-Model Prior Overcomes Cold-Start Items [14.370472820496802]
  The growth ofRecSys is driven by digitization and the need for personalized content in areas such as e-commerce and video streaming. Existing solutions for the cold-start problem, such as content-based recommenders and hybrid methods, leverage item metadata to determine item similarities. This paper introduces a novel approach for cold-start item recommendation that utilizes the language model (LM) to estimate item similarities.
  arXiv  Detail & Related papers  (2024-11-13T22:45:52Z)
• Firzen: Firing Strict Cold-Start Items with Frozen Heterogeneous and Homogeneous Graphs for Recommendation [34.414081170244955]
  We propose a unified framework incorporating multi-modal content of items and knowledge graphs (KGs) to solve both strict cold-start and warm-start recommendation. Our model yields significant improvements for strict cold-start recommendation and outperforms or matches the state-of-the-art performance in the warm-start scenario.
  arXiv  Detail & Related papers  (2024-10-10T06:48:27Z)
• Learning Multi-Aspect Item Palette: A Semantic Tokenization Framework for Generative Recommendation [55.99632509895994]
  We introduce LAMIA, a novel approach for multi-aspect semantic tokenization.<n>Unlike RQ-VAE, which uses a single embedding, LAMIA learns an item palette''--a collection of independent and semantically parallel embeddings.<n>Our results demonstrate significant improvements in recommendation accuracy over existing methods.
  arXiv  Detail & Related papers  (2024-09-11T13:49:48Z)
• Retrieval Augmentation via User Interest Clustering [57.63883506013693]
  Industrial recommender systems are sensitive to the patterns of user-item engagement. We propose a novel approach that efficiently constructs user interest and facilitates low computational cost inference. Our approach has been deployed in multiple products at Meta, facilitating short-form video related recommendation.
  arXiv  Detail & Related papers  (2024-08-07T16:35:10Z)
• Retrieval-Enhanced Contrastive Vision-Text Models [61.783728119255365]
  We propose to equip vision-text models with the ability to refine their embedding with cross-modal retrieved information from a memory at inference time. Remarkably, we show that this can be done with a light-weight, single-layer, fusion transformer on top of a frozen CLIP. Our experiments validate that our retrieval-enhanced contrastive (RECO) training improves CLIP performance substantially on several challenging fine-grained tasks.
  arXiv  Detail & Related papers  (2023-06-12T15:52:02Z)
• Diverse Preference Augmentation with Multiple Domains for Cold-start Recommendations [92.47380209981348]
  We propose a Diverse Preference Augmentation framework with multiple source domains based on meta-learning. We generate diverse ratings in a new domain of interest to handle overfitting on the case of sparse interactions. These ratings are introduced into the meta-training procedure to learn a preference meta-learner, which produces good generalization ability.
  arXiv  Detail & Related papers  (2022-04-01T10:10:50Z)
• Learning to Learn a Cold-start Sequential Recommender [70.5692886883067]
  The cold-start recommendation is an urgent problem in contemporary online applications. We propose a meta-learning based cold-start sequential recommendation framework called metaCSR. metaCSR holds the ability to learn the common patterns from regular users' behaviors.
  arXiv  Detail & Related papers  (2021-10-18T08:11:24Z)
• Sparse-Interest Network for Sequential Recommendation [78.83064567614656]
  We propose a novel textbfSparse textbfInterest textbfNEtwork (SINE) for sequential recommendation. Our sparse-interest module can adaptively infer a sparse set of concepts for each user from the large concept pool. SINE can achieve substantial improvement over state-of-the-art methods.
  arXiv  Detail & Related papers  (2021-02-18T11:03:48Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.