Neural Syntactic Preordering for Controlled Paraphrase Generation
Neural Syntactic Preordering for Controlled Paraphrase Generation
复制标题
DOI：
10.18653/v1/2020.acl-main.22
复制
发表时间：
2020-05
期刊：
影响因子：
--
通讯作者：
Tanya Goyal;Greg Durrett
中科院分区：
文献类型：
--
作者：
Tanya Goyal;Greg Durrett
释义自然语言是一个多方面的过程：它可能涉及替换单个单词或简短的短语，本地内容的重新排列，或者高级式的修复，例如局部方法或钝化方法工作灵感来自机器翻译中的预订文献，使用句法转换轻轻地“重新排序”源句子并指导我们的中性释义模型，首先，我们使用编码器模型来得出一组可行的句法重排。提议重新排列以产生一系列位置嵌入，这鼓励我们的最终编码器解码器模型以特定的顺序进行源词。人类表明，所提出的系统保留了基线方法的质量，同时使生成的释义的多样性大大增加。
Paraphrasing natural language sentences is a multifaceted process: it might involve replacing individual words or short phrases, local rearrangement of content, or high-level restructuring like topicalization or passivization. Past approaches struggle to cover this space of paraphrase possibilities in an interpretable manner. Our work, inspired by pre-ordering literature in machine translation, uses syntactic transformations to softly “reorder” the source sentence and guide our neural paraphrasing model. First, given an input sentence, we derive a set of feasible syntactic rearrangements using an encoder-decoder model. This model operates over a partially lexical, partially syntactic view of the sentence and can reorder big chunks. Next, we use each proposed rearrangement to produce a sequence of position embeddings, which encourages our final encoder-decoder paraphrase model to attend to the source words in a particular order. Our evaluation, both automatic and human, shows that the proposed system retains the quality of the baseline approaches while giving a substantial increase in the diversity of the generated paraphrases.