An Unsupervised Method for Building Sentence Simplification Corpora in Multiple Languages

Lu, Xinyu; Qiang, Jipeng; Li, Yun; Yuan, Yunhao; Zhu, Yi

doi:10.18653/v1/2021.findings-emnlp.22

Cited by 11 publications

(19 citation statements)

References 40 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this way, Martin et al (2022) employed sentence embedding modeling to measure the similarity between sentences from approximately one billion CC-NET (Wenzek et al, 2020) sentences, and subsequently constructed a new parallel SS corpora. Lu et al (2021) utilize machine translation corpus to consturct SS corpora via bach-translation technique. Compared with WikiLarge, these corpora can help to enhance the performance of supervised SS methods.…”

Section: Llmmentioning

confidence: 99%

“…In the experiments, we use standard simplification evaluation package EASSE 2 to calculate the SARI and FKGL metrics. We calculate FRES by using the script from Lu et al (2021) 3 .…”

Section: Evaluation Settingsmentioning

confidence: 99%

“…In recent years, neural SS methods utilize parallel SS datasets to train Sequence-to-Sequence models (Wang et al, 2016;Zhang and Lapata, 2017;Zhao et al, 2018) or fine-tune pretrained language models (e.g. BART (Lewis et al, 2020)) Lu et al, 2021;Martin et al, 2022). However, much work (Woodsend and Lapata, 2011;Xu et al, 2015;Qiang and Wu, 2021) pointed out that the public English SS benchmark (WikiLarge Zhang and Lapata (2017)) which align sentences from English Wikipedia and Simple English Wikipedia are deficient, because they contain a large proportion of inaccurate or inadequate simplifications, which lead to the poor generalization performance of SS methods.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Sentence Simplification via Large Language Models

Feng¹,

Qiang²,

Li³

et al. 2023

Preprint

View full text Add to dashboard Cite

Sentence Simplification aims to rephrase complex sentences into simpler sentences while retaining original meaning. Large Language models (LLMs) have demonstrated the ability to perform a variety of natural language processing tasks. However, it is not yet known whether LLMs can be served as a high-quality sentence simplification system. In this work, we empirically analyze the zero-/few-shot learning ability of LLMs by evaluating them on a number of benchmark test sets. Experimental results show LLMs outperform state-of-the-art sentence simplification methods, and are judged to be on a par with human annotators.

show abstract

Section: Llmmentioning

confidence: 99%

“…In the experiments, we use standard simplification evaluation package EASSE 2 to calculate the SARI and FKGL metrics. We calculate FRES by using the script from Lu et al (2021) 3 .…”

Section: Evaluation Settingsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Sentence Simplification via Large Language Models

Feng¹,

Qiang²,

Li³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Martin et al (2020b) proposed unsupervised mining technology to create multi-language simplification corpora automatically. Lu et al (2021) used the back-translation approach to construct a large-scale pseudo sentence simplification corpus.…”

Section: Mine Data For Simplificationmentioning

confidence: 99%

Exploiting Summarization Data to Help Text Simplification

Sun¹,

Yang²,

Wan³

2023

Preprint

View full text Add to dashboard Cite

One of the major problems with text simplification is the lack of high-quality data. The sources of simplification datasets are limited to Wikipedia and Newsela, restricting further development of this field. In this paper, we analyzed the similarity between text summarization and text simplification and exploited summarization data to help simplify. First, we proposed an alignment algorithm to extract sentence pairs from summarization datasets. Then, we designed four attributes to characterize the degree of simplification and proposed a method to filter suitable pairs. We named these pairs Sum4Simp (S4S). Next, we conducted human evaluations to show that S4S is highquality and compared it with a real simplification dataset. Finally, we conducted experiments to illustrate that the S4S can improve the performance of several mainstream simplification models, especially in low-resource scenarios.

show abstract

“…We can treat CIP task as a special paraphrase generation task. The general paraphrase generation task aims to rephrase a given sentence to another one that possesses identical semantics but various lexicons or syntax [6], [7]. Similarly, CIP emphasizes rephrasing the idioms of input sentences to word segments that reflect more intuitive and understandable paraphrasing.…”

Section: Tablementioning

confidence: 99%

Chinese Idiom Paraphrasing

Qiang¹,

Li²,

Zhang³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Idioms, are a kind of idiomatic expression in Chinese, most of which consist of four Chinese characters. Due to the properties of non-compositionality and metaphorical meaning, Chinese Idioms are hard to be understood by children and non-native speakers. This study proposes a novel task, denoted as Chinese Idiom Paraphrasing (CIP). CIP aims to rephrase idioms-included sentences to non-idiomatic ones under the premise of preserving the original sentence's meaning. Since the sentences without idioms are easier handled by Chinese NLP systems, CIP can be used to pre-process Chinese datasets, thereby facilitating and improving the performance of Chinese NLP tasks, e.g., machine translation system, Chinese idiom cloze, and Chinese idiom embeddings. In this study, CIP task is treated as a special paraphrase generation task. To circumvent difficulties in acquiring annotations, we first establish a large-scale CIP dataset based on human and machine collaboration, which consists of 115,530 sentence pairs. We further deploy three baselines and two novel CIP approaches to deal with CIP problems. The results show that the proposed methods have better performances than the baselines based on the established CIP dataset.

show abstract

An Unsupervised Method for Building Sentence Simplification Corpora in Multiple Languages

Cited by 11 publications

References 40 publications

Sentence Simplification via Large Language Models

Sentence Simplification via Large Language Models

Exploiting Summarization Data to Help Text Simplification

Chinese Idiom Paraphrasing

Contact Info

Product

Resources

About