Summary of Investigating Large Language Models and Control Mechanisms to Improve Text Readability Of Biomedical Abstracts, by Zihao Li et al.
Investigating Large Language Models and Control Mechanisms to Improve Text Readability of Biomedical Abstracts
by Zihao Li, Samuel Belkadi, Nicolo Micheletti, Lifeng Han, Matthew Shardlow, Goran Nenadic
First submitted to arxiv on: 22 Sep 2023
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 study investigates the application of state-of-the-art large language models (LLMs) for biomedical abstract simplification. The research uses a publicly available dataset, PLABA, and employs domain fine-tuning and prompt-based learning to train encoder-decoder models, decoder-only GPT models, and control-token mechanisms on BART-based models. The methods are evaluated using automatic metrics such as BLEU, ROUGE, SARI, and BERTscore, as well as human evaluations. The results show that BART-Large with Control Token (BART-L-w-CT) achieves the highest SARI score of 46.54 and T5-base reports the highest BERTscore of 72.62. The study also provides an analysis of the system outputs and sheds light on future research directions. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper explores how to simplify complex biomedical texts so that anyone can understand them. It uses special computer models called large language models to see if they can make these texts easier to read. The researchers used a big database with examples of simplified biomedical texts and tested different approaches to see which one works best. They found that some methods are better than others at making the texts simpler and more understandable. This study is important because it helps us understand how to make healthcare information more accessible to everyone. |
Keywords
» Artificial intelligence » Bleu » Decoder » Encoder decoder » Fine tuning » Gpt » Prompt » Rouge » T5 » Token
