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 2009.02835cs.CLcs.AIcs.LG
收藏

E-BERT:一种基于短语和产品知识的电子商业语言模型

E-BERT: A Phrase and Product Knowledge Enhanced Language Model for E-commerce

Rutgers University(罗格斯大学);NEC Laboratories America, Inc.(NEC美国实验室公司);Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所)

Denghui Zhang, Zixuan Yuan, Yanchi Liu, Fuzhen Zhuang, Haifeng Chen, Hui Xiong

摘要:

预先训练的语言模型,如BERT,在广泛的自然语言处理任务中取得了巨大的成功。然而,BERT由于缺乏两个层次的领域知识而不能很好地支持电子商业相关任务,即,短语级和产品级。一方面,许多电子商业任务需要对领域短语的准确理解,而这种细粒度的短语级知识并没有被BERT的训练目标明确地建模。另一方面,产品级知识,如产品关联,可以增强电子商业的语言建模,但他们不是事实知识,因此不加区别地使用它们可能会引入噪音。为了解决这个问题,我们提出了一个统一的预训练框架,即E-BERT。具体来说,为了保留短语级知识,我们引入了自适应混合掩码,它允许模型根据两种模式的拟合进度自适应地从学习初步单词知识切换到学习复杂短语。为了利用产品级知识,我们引入了Neighbor Product Reconstruction,它训练E-BERT通过去噪交叉注意层来预测产品的相关邻居。我们的调查显示,在四个下游任务,即,基于评论的问答、方面提取、方面情感分类和产品分类。

英文摘要:

Pre-trained language models such as BERT have achieved great success in a broad range of natural language processing tasks. However, BERT cannot well support E-commerce related tasks due to the lack of two levels of domain knowledge, i.e., phrase-level and product-level. On one hand, many E-commerce tasks require an accurate understanding of domain phrases, whereas such fine-grained phrase-level knowledge is not explicitly modeled by BERT's training objective. On the other hand, product-level knowledge like product associations can enhance the language modeling of E-commerce, but they are not factual knowledge thus using them indiscriminately may introduce noise. To tackle the problem, we propose a unified pre-training framework, namely, E-BERT. Specifically, to preserve phrase-level knowledge, we introduce Adaptive Hybrid Masking, which allows the model to adaptively switch from learning preliminary word knowledge to learning complex phrases, based on the fitting progress of two modes. To utilize product-level knowledge, we introduce Neighbor Product Reconstruction, which trains E-BERT to predict a product's associated neighbors with a denoising cross attention layer. Our investigation reveals promising results in four downstream tasks, i.e., review-based question answering, aspect extraction, aspect sentiment classification, and product classification.

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。

↑