New NLP data augmentation method preserves discourse structure

Researchers have developed a novel data augmentation framework for Natural Language Processing (NLP) that preserves document-level discourse structure. This method utilizes large language models and discourse parsing to identify rhetorical relations between sentences, enabling the replacement of key phrases with semantically coherent yet topically shifted content. Experiments on text summarization and question answering tasks indicate that the augmented data provides valuable complementary supervision, improving training performance when combined with existing data. AI

IMPACT Enhances NLP model performance in low-resource settings by improving data augmentation techniques.

RANK_REASON The item is an academic paper detailing a new method for data augmentation in NLP. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• alphaXiv
• arXiv
• CatalyzeX Code Finder for Papers
• Connected Papers
• CORE Recommender
• DagsHub
• Gotit.pub
• Hugging Face
• large language models
• Litmaps
• Natural Language Processing
• question answering
• RST parser
• ScienceCast
• scite Smart Citations

• paper
• product

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →