Dataset: Automated Quality Assessment of Online Medical Science Popularization Videos

This repository contains all the necessary data files for the readers of “Automated Quality Assessment of Online Medical Science Popularization Videos: A Framework Based on Large Language Models.”
· Converted_Text_with_Human_Ratings.xlsx: Deidentified text transcripts converted from video audio tracks with corresponding human ratings.
· likelihood_selection.py: The code to realize automated quality assessment of videos with large language models using the likelihood selection strategy.
· LLM_Evaluation_Results: Evaluation results of 13 large language models.
· Resilience_to_Distrubance: Evaluation results of LLaMA-3 based models with different prompt generation strategies as disturbance.