Scholarlink

<p>Sequential recommendation aims to predict future interactions by modeling historical user behavior, traditionally utilizing stacked Transformer architectures to capture interest evolution. However, pervasive data sparsity often triggers representation degradation, causing deep-layer features to homogenize, which impairs both expressive capability and optimization efficiency. Although data augmentation mitigates sparse signals, it frequently compromises semantic consistency across layers, undermining model robustness. Moreover, existing models rarely account for varying sparsity levels across datasets, which demand adaptive structural complexity. To address these challenges, we propose Distilled and Sparsity Adaptive Mamba for Sequential Recommendation (DiSAM4Rec). Leveraging Mamba’s linear complexity, we first design a Mamba-based Sparsity-Aware Dual-Branch (MSADB) module that dynamically switches between dense and sparse branches to optimize feature extraction across varying interaction densities. Based on this, we introduce a Layer-wise Semantic Alignment Self-Distillation (LSA-SD) mechanism to align intermediate features with high-level abstract semantic spaces, thereby mitigating representation degradation and ensuring deep-layer consistency. Extensive experiments on four real-world datasets demonstrate that DiSAM4Rec significantly outperforms competitive benchmarks, achieving average improvements of 4.12%, 9.24%, and 9.97% across HIT@10, NDCG@10, and MRR@10, while maintaining lower computational overhead. The code is available at <a href="https://github.com/HammerLau/DiSAM4Rec">https://github.com/HammerLau/DiSAM4Rec</a>.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

DiSAM4Rec: distilled and sparsity adaptive Mamba for sequential recommendation

  • 期刊论文

  • 发表日期：

    2026年7月3日

查看全文
  • Ruiming Liu,
  • Zhi Liu,
  • Xincheng Xia,
  • Qing Qing

摘要

Sequential recommendation aims to predict future interactions by modeling historical user behavior, traditionally utilizing stacked Transformer architectures to capture interest evolution. However, pervasive data sparsity often triggers representation degradation, causing deep-layer features to homogenize, which impairs both expressive capability and optimization efficiency. Although data augmentation mitigates sparse signals, it frequently compromises semantic consistency across layers, undermining model robustness. Moreover, existing models rarely account for varying sparsity levels across datasets, which demand adaptive structural complexity. To address these challenges, we propose Distilled and Sparsity Adaptive Mamba for Sequential Recommendation (DiSAM4Rec). Leveraging Mamba’s linear complexity, we first design a Mamba-based Sparsity-Aware Dual-Branch (MSADB) module that dynamically switches between dense and sparse branches to optimize feature extraction across varying interaction densities. Based on this, we introduce a Layer-wise Semantic Alignment Self-Distillation (LSA-SD) mechanism to align intermediate features with high-level abstract semantic spaces, thereby mitigating representation degradation and ensuring deep-layer consistency. Extensive experiments on four real-world datasets demonstrate that DiSAM4Rec significantly outperforms competitive benchmarks, achieving average improvements of 4.12%, 9.24%, and 9.97% across HIT@10, NDCG@10, and MRR@10, while maintaining lower computational overhead. The code is available at https://github.com/HammerLau/DiSAM4Rec.