Large-Scale, Diverse, Paraphrastic Bitexts via Sampling and Clustering

By J. Edward Hu, Abhinav Singh, Nils Holzenberger, Matt Post, Benjamin Van Durme. 2019

70 citations recorded by the source corpus.

Producing diverse paraphrases of a sentence is a challenging task. Natural paraphrase corpora are scarce and limited, while existing large-scale resources are automatically generated via back-translation and rely on beam search, which tends to lack diversity. We describe ParaBank 2, a new resource that contains multiple diverse sentential paraphrases, produced from a bilingual corpus using negative constraints, inference sampling, and clustering.We show that ParaBank 2 significantly surpasses prior work in both lexical and syntactic diversity while being meaning-preserving, as measured by human judgments and standardized metrics. Further, we illustrate how such paraphrastic resources may be used to refine contextualized encoders, leading to improvements in downstream tasks.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Machine Translation

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Machine Translation

Citations in the corpus

GPTScore: Evaluate as You Desire (2023, neutral): “0 ( Hu et al., 2019).”
BARTScore: Evaluating Generated Text as Text Generation (2021, methodology): “We continue ﬁne-tuning BART from (1) on ParaBank2 dataset [22], which contains a large paraphrase collection.”
AutoQA: From Databases to Q&A Semantic Parsers with Only Synthetic Training Data (2020, methodology): “We fine-tune it for 4 epochs on sentence pairs from PARABANK 2 (Hu et al., 2019a), which is a paraphrase dataset constructed by back-translating the Czech portion of an English-Czech parallel corpus.”
Impossible Distillation for Paraphrasing and Summarization: How to Make High-quality Lemonade out of Small, Low-quality Model (2023, methodology): “Notably, the high faithfulness and dissimilarity does not come from sacrificing the soundness of generation, marking better fluency score than both T5 ParaBank and T5 ChatGPT-Para .”
TRUE: Re-evaluating Factual Consistency Evaluation (2022, methodology): “We use the version ﬁne-tuned on the ParaBank2 dataset (Hu et al., 2019).”
COD3S: Diverse Generation with Discrete Semantic Signatures (2020, methodology): “Recent methods for improved diversity generation make slight modiﬁcations to the neural architecture or beam search algorithm (Xu et al., 2018; Li et al., 2016b), or impose lexical constraints during decoding (Post and Vilar, 2018; Hu et al., 2019a).”