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 1808.09042cs.CL
收藏

文本表示的对抗性分解

Adversarial Decomposition of Text Representation

University of Massachusetts Lowell(马萨诸塞大学洛厄尔分校)

Alexey Romanov, Anna Rumshisky, Anna Rogers, David Donahue

摘要:

在本文中,我们提出了一种方法的对抗分解的文本表示。这种方法可以用来将一个输入句子的表示分解成几个独立的向量,每个向量负责输入句子的一个特定方面。我们评估两个案例研究所提出的方法:不同的社会寄存器和历时语言变化之间的转换。我们表明,所提出的方法是能够细粒度控制的变化,这些方面的输入句子。它也在学习句子风格的连续(而不是分类)表示,这在语言学上更现实。该模型使用对抗激励训练,并包括一个特殊的激励损失,它的作用与对抗激励相反,并鼓励更好的分解。此外,我们评估所获得的意义嵌入的下游任务的释义检测,并表明它们显着优于嵌入的一个定期的自动编码器。

英文摘要:

In this paper, we present a method for adversarial decomposition of text representation. This method can be used to decompose a representation of an input sentence into several independent vectors, each of them responsible for a specific aspect of the input sentence. We evaluate the proposed method on two case studies: the conversion between different social registers and diachronic language change. We show that the proposed method is capable of fine-grained controlled change of these aspects of the input sentence. It is also learning a continuous (rather than categorical) representation of the style of the sentence, which is more linguistically realistic. The model uses adversarial-motivational training and includes a special motivational loss, which acts opposite to the discriminator and encourages a better decomposition. Furthermore, we evaluate the obtained meaning embeddings on a downstream task of paraphrase detection and show that they significantly outperform the embeddings of a regular autoencoder.

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。

↑