arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2102.00467cs.CLcs.LG
收藏

混合正则化对抗网络在多领域文本分类中的应用

Mixup Regularized Adversarial Networks for Multi-Domain Text Classification

Carleton University(卡尔顿大学);University of Ottawa(渥太华大学)

Yuan Wu, Diana Inkpen, Ahmed El-Roby

摘要:

使用共享私有范式和对抗训练显著提高了多域文本分类(MDTC)模型的性能。然而,现有方法存在两个问题。首先,来自多个域的实例对于域不变的特征提取是不够的。其次,在边缘分布上对齐可能导致致命的失配。在本文中,我们提出了一个混合正则化对抗网络(MRAN)来解决这两个问题。更具体地说,引入域和类别混合正则化来丰富共享潜在空间中的内在特征,并在训练实例之间执行一致的预测,使得学习的特征可以更具域不变性和区分性。我们在两个基准上进行实验:亚马逊评论数据集和FDU-MTL数据集。我们在这两个数据集上的方法分别产生了87.64%和89.0%的平均准确率,优于所有相关基线。

英文摘要:

Using the shared-private paradigm and adversarial training has significantly improved the performances of multi-domain text classification (MDTC) models. However, there are two issues for the existing methods. First, instances from the multiple domains are not sufficient for domain-invariant feature extraction. Second, aligning on the marginal distributions may lead to fatal mismatching. In this paper, we propose a mixup regularized adversarial network (MRAN) to address these two issues. More specifically, the domain and category mixup regularizations are introduced to enrich the intrinsic features in the shared latent space and enforce consistent predictions in-between training instances such that the learned features can be more domain-invariant and discriminative. We conduct experiments on two benchmarks: The Amazon review dataset and the FDU-MTL dataset. Our approach on these two datasets yields average accuracies of 87.64\% and 89.0\% respectively, outperforming all relevant baselines.

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。

↑