Reproduce “Neural Automated Writing Evaluation with Corrective Feedback” 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
Izia Xiaoxiao Wang, Xihan Wu, Edith Coates, Min Zeng, Jiexin Kuang, Siliang Liu, Mengyang Qiu, Jungyeul Park
Researchers at The University of British Columbia developed a neural system that combines Automated Writing Evaluation (AWE) and Grammatical Error Correction (GEC), providing L2 learners with instant, objective writing scores and in-line grammatical feedback, achieving enhanced AWE accuracy via text denoising and competitive GEC performance.
Key insights
- Meticulous text denoising of L2 writing datasets significantly enhances the accuracy of Automated Writing Evaluation models, improving linear regression results for essay scoring.
- Neural AWE and GEC systems can be effectively integrated to deliver a comprehensive feedback mechanism that addresses both writing quality assessment and grammatical error correction.
- Such an integrated system allows L2 learners to practice essay writing in a simulated test environment, receiving immediate, objective scores and corrective feedback.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.