DeepQR: Neural-Based Quality Ratings for Learnersourced Multiple-Choice Questions

Automated question quality rating (AQQR) aims to evaluate through computational means, thereby addressing emerging challenges in online learnersourced repositories. Existing methods for AQQR rely solely on explicitly-defined criteria such as readability and word count, while not fully utilising the power of state-of-the-art deep-learning techniques. We propose DeepQR, a novel neural-network model that is trained using multiple-choice-question (MCQ) datasets collected from PeerWise, widely-used learnersourcing platform. Along with designing we investigate models based features, or semantic both. also introduce self-attention mechanism capture correlations between MCQ components, contrastive-learning approach acquire representations ratings. Extensive experiments eight university-level courses illustrate DeepQR has superior performance over six comparative models.