English

Learning Effective and Efficient Embedding via an Adaptively-Masked Twins-based Layer

Machine Learning 2021-08-27 v1 Artificial Intelligence

Abstract

Embedding learning for categorical features is crucial for the deep learning-based recommendation models (DLRMs). Each feature value is mapped to an embedding vector via an embedding learning process. Conventional methods configure a fixed and uniform embedding size to all feature values from the same feature field. However, such a configuration is not only sub-optimal for embedding learning but also memory costly. Existing methods that attempt to resolve these problems, either rule-based or neural architecture search (NAS)-based, need extensive efforts on the human design or network training. They are also not flexible in embedding size selection or in warm-start-based applications. In this paper, we propose a novel and effective embedding size selection scheme. Specifically, we design an Adaptively-Masked Twins-based Layer (AMTL) behind the standard embedding layer. AMTL generates a mask vector to mask the undesired dimensions for each embedding vector. The mask vector brings flexibility in selecting the dimensions and the proposed layer can be easily added to either untrained or trained DLRMs. Extensive experimental evaluations show that the proposed scheme outperforms competitive baselines on all the benchmark tasks, and is also memory-efficient, saving 60\% memory usage without compromising any performance metrics.

Keywords

Cite

@article{arxiv.2108.11513,
  title  = {Learning Effective and Efficient Embedding via an Adaptively-Masked Twins-based Layer},
  author = {Bencheng Yan and Pengjie Wang and Kai Zhang and Wei Lin and Kuang-Chih Lee and Jian Xu and Bo Zheng},
  journal= {arXiv preprint arXiv:2108.11513},
  year   = {2021}
}

Comments

CIKM 2021, 5 pages; The first two authors contributed equally to this work

R2 v1 2026-06-24T05:25:34.285Z