Embedding Enhancement via Fine-Tuned Language Models for Learner-Item Cognitive Modeling

2026-04-05

AI Summary

This work addresses the semantic misalignment between pre-trained language models and cognitive diagnostic models, which arises from divergent training objectives and the absence of a unified framework to integrate textual semantics for diverse diagnostic tasks. To bridge this gap, the authors propose EduEmbed, a two-stage framework that first fine-tunes the language model using role-based representations and an interaction-aware diagnostic head to align semantic spaces, then introduces a text adapter to extract task-relevant semantics and seamlessly integrate them with established cognitive diagnosis paradigms. EduEmbed is the first approach to enable embedding-augmented unified modeling, preserving the strengths of traditional diagnostic methods while substantially enhancing cross-task generalization. Empirical results demonstrate consistent performance gains across four cognitive diagnosis tasks and computerized adaptive testing, confirming the effective enrichment of educational intelligence through semantic information.

Abstract

Learner-item cognitive modeling plays a central role in the web-based online intelligent education system by enabling cognitive diagnosis (CD) across diverse online educational scenarios. Although ID embedding remains the mainstream approach in cognitive modeling due to its effectiveness and flexibility, recent advances in language models (LMs) have introduced new possibilities for incorporating rich semantic representations to enhance CD performance. This highlights the need for a comprehensive analysis of how LMs enhance embeddings through semantic integration across mainstream CD tasks. This paper identifies two key challenges in fully leveraging LMs in existing work: Misalignment between the training objectives of LMs and CD models creates a distribution gap in feature spaces; A unified framework is essential for integrating textual embeddings across varied CD tasks while preserving the strengths of existing cognitive modeling paradigms to ensure the robustness of embedding enhancement. To address these challenges, this paper introduces EduEmbed, a unified embedding enhancement framework that leverages fine-tuned LMs to enrich learner-item cognitive modeling across diverse CD tasks. EduEmbed operates in two stages. In the first stage, we fine-tune LMs based on role-specific representations and an interaction diagnoser to bridge the semantic gap of CD models. In the second stage, we employ a textual adapter to extract task-relevant semantics and integrate them with existing modeling paradigms to improve generalization. We evaluate the proposed framework on four CD tasks and computerized adaptive testing (CAT) task, achieving robust performance. Further analysis reveals the impact of semantic information across diverse tasks, offering key insights for future research on the application of LMs in CD for online intelligent education systems.

Problem

• cognitive diagnosis
• language models
• embedding enhancement
• learner-item modeling
• semantic integration

Innovation

• embedding enhancement
• fine-tuned language models
• cognitive diagnosis
• textual adapter
• semantic integration