Unsupervised Paraphrasing via Deep Reinforcement Learning
Unsupervised Paraphrasing via Deep Reinforcement Learning
复制标题
DOI：
10.1145/3394486.3403231
复制
发表时间：
2020-07
期刊：
影响因子：
--
通讯作者：
A.B. Siddique;Samet Oymak;Vagelis Hristidis
中科院分区：
文献类型：
--
作者：
A.B. Siddique;Samet Oymak;Vagelis Hristidis
释义是用不同的措辞表达输入句子的含义，同时保持流畅性（即，语法和句法正确性）。大多数关于释义的现有工作使用限于特定领域的监督模型（例如，图像字幕）。这样的模型既不能直接转移到其他领域，也不能很好地推广，而且为新领域创建标记的训练数据既昂贵又费力。跨领域释义的需求以及许多此类领域中标记训练数据的稀缺性要求探索无监督释义生成方法。我们提出了渐进式无监督释义（PUP）：一种基于深度强化学习（DRL）的新型无监督释义生成方法。PUP使用变分自动编码器（使用非并行语料库进行训练）来生成种子释义，从而热启动DRL模型。然后，PUP在我们的新颖奖励函数的指导下逐步调整种子释义，该奖励函数结合了语义充分性，语言流畅性和表达多样性措施，以量化每次迭代中生成的释义的质量，而无需并行句子。我们广泛的实验评估表明，PUP优于无监督的国家的最先进的释义技术在自动度量和用户研究四个真实的数据集。我们还表明，PUP在几个数据集上的性能优于域适应监督算法。我们的评估还表明，PUP实现了语义相似性和表达的多样性之间的很大的权衡。
Paraphrasing is expressing the meaning of an input sentence in different wording while maintaining fluency (i.e., grammatical and syntactical correctness). Most existing work on paraphrasing use supervised models that are limited to specific domains (e.g., image captions). Such models can neither be straightforwardly transferred to other domains nor generalize well, and creating labeled training data for new domains is expensive and laborious. The need for paraphrasing across different domains and the scarcity of labeled training data in many such domains call for exploring unsupervised paraphrase generation methods. We propose Progressive Unsupervised Paraphrasing (PUP): a novel unsupervised paraphrase generation method based on deep reinforcement learning (DRL). PUP uses a variational autoencoder (trained using a non-parallel corpus) to generate a seed paraphrase that warm-starts the DRL model. Then, PUP progressively tunes the seed paraphrase guided by our novel reward function which combines semantic adequacy, language fluency, and expression diversity measures to quantify the quality of the generated paraphrases in each iteration without needing parallel sentences. Our extensive experimental evaluation shows that PUP outperforms unsupervised state-of-the-art paraphrasing techniques in terms of both automatic metrics and user studies on four real datasets. We also show that PUP outperforms domain-adapted supervised algorithms on several datasets. Our evaluation also shows that PUP achieves a great trade-off between semantic similarity and diversity of expression.