arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2004.13139cs.IRcs.LG
收藏

一种面向顺序推荐系统的通用网络压缩框架

A Generic Network Compression Framework for Sequential Recommender Systems

University of Science and Technology of China(中国科学技术大学);Tencent(腾讯);SIAT, Chinese Academy of Sciences(中国科学院深圳先进技术研究院);Zhejiang University(浙江大学)

Yang Sun, Fajie Yuan, Min Yang, Guoao Wei, Zhou Zhao, Duo Liu

摘要:

顺序推荐系统(SRS)已经成为捕捉用户动态兴趣和生成高质量推荐的关键技术。当前最先进的顺序推荐模型通常基于嵌入式结构的深度神经网络,其中一个或多个中间(隐藏)层放置在输入嵌入层和输出softmax层之间。一般来说,这些模型需要大量的参数(例如使用大的嵌入维度或深度网络架构)来获得其最佳性能。尽管有效,但在某些时候,进一步增加模型大小可能更难在资源受限设备中部署模型,导致更长的响应时间和更大的内存占用。为了解决这些问题,我们提出了一个压缩的顺序推荐框架,称为CpRec,其中两个通用的模型收缩技术。具体来说,我们首先提出了一个块式自适应分解近似的输入和softmax矩阵,利用SRS中的项目服从长尾分布的事实。为了减少中间层的参数,我们引入了三个逐层参数共享计划。我们使用深度卷积神经网络实例化CpRec,考虑到推荐准确性和效率,使用膨胀的内核。通过广泛的消融研究,我们证明了所提出的CpRec可以在真实世界的SRS数据集上实现高达4 $\sim$ 8倍的压缩率。同时,CpRec在训练\inference期间更快,并且在大多数情况下优于其未压缩的对应物。

英文摘要:

Sequential recommender systems (SRS) have become the key technology in capturing user's dynamic interests and generating high-quality recommendations. Current state-of-the-art sequential recommender models are typically based on a sandwich-structured deep neural network, where one or more middle (hidden) layers are placed between the input embedding layer and output softmax layer. In general, these models require a large number of parameters (such as using a large embedding dimension or a deep network architecture) to obtain their optimal performance. Despite the effectiveness, at some point, further increasing model size may be harder for model deployment in resource-constraint devices, resulting in longer responding time and larger memory footprint. To resolve the issues, we propose a compressed sequential recommendation framework, termed as CpRec, where two generic model shrinking techniques are employed. Specifically, we first propose a block-wise adaptive decomposition to approximate the input and softmax matrices by exploiting the fact that items in SRS obey a long-tailed distribution. To reduce the parameters of the middle layers, we introduce three layer-wise parameter sharing schemes. We instantiate CpRec using deep convolutional neural network with dilated kernels given consideration to both recommendation accuracy and efficiency. By the extensive ablation studies, we demonstrate that the proposed CpRec can achieve up to 4$\sim$8 times compression rates in real-world SRS datasets. Meanwhile, CpRec is faster during training\inference, and in most cases outperforms its uncompressed counterpart.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑