CENTAL at TSAR-2022 Shared Task: How Does Context Impact BERT-Generated Substitutions for Lexical Simplification?

Proceedings of the Workshop on Text Simplification, Accessibility, and Readability (TSAR-2022) Pub Date : 1900-01-01 DOI:10.18653/v1/2022.tsar-1.25

Rodrigo Wilkens, David Alfter, Rémi Cardon, Isabelle Gribomont, Adrien Bibal, Watrin Patrick, Marie-Catherine de Marneffe, Thomas François

{"title":"CENTAL at TSAR-2022 Shared Task: How Does Context Impact BERT-Generated Substitutions for Lexical Simplification?","authors":"Rodrigo Wilkens, David Alfter, Rémi Cardon, Isabelle Gribomont, Adrien Bibal, Watrin Patrick, Marie-Catherine de Marneffe, Thomas François","doi":"10.18653/v1/2022.tsar-1.25","DOIUrl":null,"url":null,"abstract":"Lexical simplification is the task of substituting a difficult word with a simpler equivalent for a target audience. This is currently commonly done by modeling lexical complexity on a continuous scale to identify simpler alternatives to difficult words. In the TSAR shared task, the organizers call for systems capable of generating substitutions in a zero-shot-task context, for English, Spanish and Portuguese. In this paper, we present the solution we (the {textsc{cental} team) proposed for the task. We explore the ability of BERT-like models to generate substitution words by masking the difficult word. To do so, we investigate various context enhancement strategies, that we combined into an ensemble method. We also explore different substitution ranking methods. We report on a post-submission analysis of the results and present our insights for potential improvements. The code for all our experiments is available at https://gitlab.com/Cental-FR/cental-tsar2022.","PeriodicalId":247582,"journal":{"name":"Proceedings of the Workshop on Text Simplification, Accessibility, and Readability (TSAR-2022)","volume":"28 1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1900-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the Workshop on Text Simplification, Accessibility, and Readability (TSAR-2022)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.18653/v1/2022.tsar-1.25","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}

引用次数: 6

Abstract

Lexical simplification is the task of substituting a difficult word with a simpler equivalent for a target audience. This is currently commonly done by modeling lexical complexity on a continuous scale to identify simpler alternatives to difficult words. In the TSAR shared task, the organizers call for systems capable of generating substitutions in a zero-shot-task context, for English, Spanish and Portuguese. In this paper, we present the solution we (the {textsc{cental} team) proposed for the task. We explore the ability of BERT-like models to generate substitution words by masking the difficult word. To do so, we investigate various context enhancement strategies, that we combined into an ensemble method. We also explore different substitution ranking methods. We report on a post-submission analysis of the results and present our insights for potential improvements. The code for all our experiments is available at https://gitlab.com/Cental-FR/cental-tsar2022.

查看原文

微信好友朋友圈 QQ好友复制链接

本刊更多论文

TSAR-2022共享任务:上下文如何影响bert生成的词汇简化替代?

词汇简化是指为目标受众用较简单的对应词代替较难的词。目前，这通常是通过在连续尺度上对词汇复杂性进行建模来识别难词的简单替代词来完成的。在TSAR共享任务中，组织者要求系统能够在零射门任务上下文中生成替换，用于英语，西班牙语和葡萄牙语。在本文中，我们提出了我们({textsc{central}团队)为该任务提出的解决方案。我们探索了类bert模型通过屏蔽难词来生成替代词的能力。为此，我们研究了各种上下文增强策略，并将其组合成一个集成方法。我们还探讨了不同的替代排序方法。我们报告提交后的结果分析，并提出我们对潜在改进的见解。我们所有实验的代码都可以在https://gitlab.com/Cental-FR/cental-tsar2022上找到。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文去求助

来源期刊

Proceedings of the Workshop on Text Simplification, Accessibility, and Readability (TSAR-2022)

自引率

0.00%

发文量