Researchers have developed a new framework called Process-aware Language Cognitive Diagnosis (PLCD) that utilizes large language models (LLMs) to improve cognitive diagnosis in online learning. Unlike traditional methods that use discrete student IDs, PLCD incorporates language-derived structures and response records to better represent student knowledge and predict performance. The framework constructs concept schemas and cognitive process graphs, and uses semantic memory to retrieve relevant historical responses. Experiments indicate that PLCD surpasses existing baselines and demonstrates strong cognitive transfer capabilities, suggesting that LLMs can enhance the measurement of latent knowledge states. AI
IMPACT This research could lead to more personalized and effective online learning experiences by improving how student knowledge is assessed.
RANK_REASON Academic paper detailing a new framework for cognitive diagnosis using LLMs. [lever_c_demoted from research: ic=1 ai=1.0]
- arXiv
- Cognitive diagnosis models for multiple strategies
- DA-MoE
- Language-to-Cognition Mapper
- LLMs
- Process-aware Language Cognitive Diagnosis
- Process-Grounded Language Modeling for Cognitive Diagnosis
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
