Simple is not Enough: Document-level Text Simplification using Readability and Coherence

Laura Vásquez-Rodríguez, Nhung T.H. Nguyen, Piotr Przybyła, Matthew Shardlow, Sophia Ananiadou · 2024

Read Simple is not Enough: Document-level Text Simplification using Readability and Coherence by Laura Vásquez-Rodríguez, Nhung T.H. Nguyen, Piotr Przybyła, Matthew Shardlow, Sophia Ananiadou (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-18655.txt.

Description

Laura Vásquez-Rodríguez and co-authors present SimDoc in 2024, a system that simplifies whole documents, not just single sentences. Simple texts help many readers, yet most tools work line by line and lose the thread of a passage. SimDoc weighs three things at once: simplicity, readability, and coherence.

The system first trains on texts rewritten by professionals. The authors extended that data with readability labels so the model learns what easy reading looks like. They then compared training styles, from no examples to full fine-tuning, on document-level sets. Their model kept the meaning while easing the load. A close look at outputs shows where document-level simplification still struggles.

Catalog record

CREATORLaura Vásquez-Rodríguez, Nhung T.H. Nguyen, Piotr Przybyła, Matthew Shardlow, Sophia Ananiadou
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE52 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-18655

Browse the full library · About the Tower of Babel