[论文解读] How to Retrain Recommender System? A Sequential Meta-Learning Method

Yang Zhang, Fuli Feng+5 more|arXiv (Cornell University)|May 27, 2020

Recommender Systems and Techniques参考文献 46被引用 8

一句话总结

本文提出顺序元学习（SML），一种通过学习迁移网络将旧模型适配到新数据的推荐系统微调方法，无需重新训练历史数据。SML 以未来推荐性能为目标进行优化，在两个真实世界数据集上实现了比完整微调更快的训练速度和更高的准确率。

ABSTRACT

Practical recommender systems need be periodically retrained to refresh the model with new interaction data. To pursue high model fidelity, it is usually desirable to retrain the model on both historical and new data, since it can account for both long-term and short-term user preference. However, a full model retraining could be very time-consuming and memory-costly, especially when the scale of historical data is large. In this work, we study the model retraining mechanism for recommender systems, a topic of high practical values but has been relatively little explored in the research community. Our first belief is that retraining the model on historical data is unnecessary, since the model has been trained on it before. Nevertheless, normal training on new data only may easily cause overfitting and forgetting issues, since the new data is of a smaller scale and contains fewer information on long-term user preference. To address this dilemma, we propose a new training method, aiming to abandon the historical data during retraining through learning to transfer the past training experience. Specifically, we design a neural network-based transfer component, which transforms the old model to a new model that is tailored for future recommendations. To learn the transfer component well, we optimize the "future performance" -- i.e., the recommendation accuracy evaluated in the next time period. Our Sequential Meta-Learning(SML) method offers a general training paradigm that is applicable to any differentiable model. We demonstrate SML on matrix factorization and conduct experiments on two real-world datasets. Empirical results show that SML not only achieves significant speed-up, but also outperforms the full model retraining in recommendation accuracy, validating the effectiveness of our proposals. We release our codes at: https://github.com/zyang1580/SML.

研究动机与目标

• 解决大规模推荐系统中完整模型微调的低效与高成本问题。
• 通过在不访问历史数据的情况下保留长期用户偏好，克服微调中的过拟合与灾难性遗忘问题。
• 明确针对未来推荐性能优化微调过程，而非依赖启发式数据采样或人工调参。
• 开发一种通用的、与模型无关的训练范式，适用于可微分推荐模型。

提出的方法

• 提出一种顺序元学习框架，通过新交互数据学习一个迁移网络，将旧模型映射到新模型。
• 利用未来验证数据（下一个时间周期的交互）来优化迁移网络，直接针对未来数据的推荐准确率进行优化。
• 采用元优化目标，最小化未来数据上的期望损失，使模型能够很好地泛化到未来的用户偏好。
• 设计基于神经网络的迁移组件，学习在不重新处理历史数据的情况下，将旧模型参数适配到新数据分布。
• 将该方法应用于矩阵分解模型，证明其在标准协同过滤架构中的适用性。
• 将每次微调周期视为一个元任务，使模型能够从连续数据周期中学习泛化能力。

实验结果

研究问题

• RQ1 我们能否在不重新处理大规模历史数据集的情况下，高效地微调推荐模型？
• RQ2 在适应新短期偏好时，如何保留长期用户偏好信号？
• RQ3 能否将未来推荐性能直接用作模型微调的优化目标？
• RQ4 学习迁移过往训练经验是否能在速度和准确率上均优于完整微调？

主要发现

• 与完整模型微调相比，SML 显著提升了训练速度，因为它在每次微调周期中避免了对历史数据的处理。
• 在两个真实世界数据集上，SML 在推荐准确率上优于完整微调，展现出对未来发展数据更强的泛化能力。
• SML 有效缓解了过拟合与灾难性遗忘问题，即使仅在新数据上训练，也能通过利用从过往模型中学习到的迁移知识实现良好表现。
• 该方法在不访问历史交互数据的情况下保持了高模型保真度，验证了‘过往训练知识可被有效迁移’的假设。
• 迁移网络学习到了有意义的表征，通过滤波器分析可视化显示其对新数据模式的有效适应。