Is Sliding Window All You Need? An Open Framework for Long-Sequence Recommendation

2026-04-14

AI Summary

This work addresses the challenge of deploying long-sequence recommendation models under memory and latency constraints in resource-limited settings. The authors propose an end-to-end open-source framework that leverages a sliding window mechanism for efficient training and introduces a k-shift embedding layer, enabling support for million-scale vocabularies on commodity GPUs with negligible accuracy degradation. For the first time, they conduct a systematic, runtime-aware ablation study of windowing strategies, thoroughly evaluating the trade-offs between recommendation accuracy and computational overhead. Experiments on datasets such as RetailRocket demonstrate substantial performance gains—up to +6.04% in MRR and +6.34% in Recall@10—with only approximately 4× increased training cost, allowing stable deployment on university-scale GPU clusters.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Machine Learning: Large Multimodal Models (LMMs)
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Long interaction histories are central to modern recommender systems, yet training with long sequences is often dismissed as impractical under realistic memory and latency budgets. This work demonstrates that it is not only practical but also effective-at academic scale. We release a complete, end-to-end framework that implements industrial-style long-sequence training with sliding windows, including all data processing, training, and evaluation scripts. Beyond reproducing prior gains, we contribute two capabilities missing from earlier reports: (i) a runtime-aware ablation study that quantifies the accuracy-compute frontier across windowing regimes and strides, and (ii) a novel k-shift embedding layer that enables million-scale vocabularies on commodity GPUs with negligible accuracy loss. Our implementation trains reliably on modest university clusters while delivering competitive retrieval quality (e.g., up to +6.04% MRR and +6.34% Recall@10 on Retailrocket) with $\sim 4 \times $ training-time overheads. By packaging a robust pipeline, reporting training time costs, and introducing an embedding mechanism tailored for low-resource settings, we transform long-sequence training from a closed, industrial technique into a practical, open, and extensible methodology for the community.

Problem

• long-sequence recommendation
• sliding window
• training efficiency
• memory constraints
• recommendation systems

Innovation

• sliding window
• long-sequence recommendation
• k-shift embedding
• runtime-aware ablation
• open framework