Summary of Multils: a Multi-task Lexical Simplification Framework, by Kai North et al.
MultiLS: A Multi-task Lexical Simplification Framework
by Kai North, Tharindu Ranasinghe, Matthew Shardlow, Marcos Zampieri
First submitted to arxiv on: 22 Feb 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper presents MultiLS, a novel framework for creating multi-task lexical simplification (LS) datasets. LS aims to improve text accessibility by replacing difficult words with easier alternatives while preserving the original meaning. The existing LS datasets specialize in one or two sub-tasks within the pipeline, but none cover all tasks. The authors introduce MultiLS-PT, the first dataset created using the MultiLS framework, which can perform lexical complexity prediction (LCP), substitute generation, and ranking for Portuguese. Model performances are reported, ranging from transformer-based models to large language models (LLMs). This research contributes to improving text accessibility for various target demographics, including children, second language learners, individuals with reading disabilities or low literacy. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about making complicated words easier to read and understand. They want to help people who have trouble reading complex texts, like kids learning English as a second language or people with reading difficulties. Right now, there are some datasets that can do part of this task, but none that can do all the steps. The authors created a new way called MultiLS, which can simplify text in three ways: guessing how hard the words are to read, coming up with simpler words, and choosing the best one. They tested it on Portuguese texts and showed that their method works well. |
Keywords
» Artificial intelligence » Multi task » Transformer
