Reproduce “Matching Exemplar as Next Sentence Prediction (MeNSP): Zero-shot Prompt Learning for Automatic Scoring in Science Education” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Xuansheng Wu, Xinyu He, Tianming Liu, Ninghao Liu, Xiaoming Zhai
MeNSP uses a pretrained BERT model to score science arguments by matching student responses with expert-written exemplars for each score level, rather than training on student answers. It assigns the level with the strongest next-sentence-prediction match. On G5, zero-shot scoring reached Cohen’s κ of 57.2% and weighted F1 of 81.1%, outperforming random scoring; evaluation covered only three related gas-context tasks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.