摘要
近年来,文本风格转换作为一种可控的文本生成任务受到学者们越来越多的关注。该文基于变分自编码器模型,通过鉴别器与变分自编码器的对抗性训练,将源端句子的内容和风格在隐变量空间进行分离,从而实现无监督的文本风格转换。针对文本语义内容和风格的解纠缠过程中利用固定的二进制向量通过线性变换来对风格进行表征的方法的不足,该文提出更具细腻度的联合表征方法:利用独立的编码器从原句中提取风格的连续隐向量,再和标签向量结合作为最终风格的表征,以提升风格转换的准确率。该文提出的联合表征方法在常用数据集Yelp上进行评测,与两个基线方法相比,风格转换准确率均有显著提升。
As a controllable text generation task,text style transfer has attracted more and more attention in recent years.Based on the variational auto-encoder model,the content and style of source sentences are separated in the latent space through the adversarial training between the discriminator and the variational auto-encoder.Due to the defect in the method using fixed binary vector for the style representation,we proposed a more fine-grained joint representation method which combines the latent variable extracted from an independent encoder with a style label to improve the accuracy of style transferation.Experimental results show that the joint representation method achieves higher accuracy compared with two baseline models on Yelp,a common dataset in the style transfer field.
作者
聂锦燃
魏蛟龙
唐祖平
NIE Jinran;WEI Jiaolong;TANG Zuping(School of Electronic Information and Communications,Huazhong University of Science and Technology,Wuhan,Hubei 430074,China)
出处
《中文信息学报》
CSCD
北大核心
2020年第7期79-88,共10页
Journal of Chinese Information Processing
基金
国家重点研发计划(2018YFB1004600)
关键词
文本风格转换
变分自编码器
对抗性训练
联合表征
text style transfer
variational auto-encoder
adversarial training
joint representation
