Research map / Large language models / Retrieval and knowledge
HISE-KT: Synergizing Heterogeneous Information Networks and LLMs for Explainable Knowledge Tracing with Meta-Path Optimization
Zhiyi Duan, Zixing Shi, Hongyu Yuan, Qi Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add similar-student retrieval context to the LLM prompt (ablation) | HISE-KT (Qwen-Plus) · Slepemapy | AUC: 0.6911 -> 0.9749 | Improved |
| add similar-student retrieval context to the LLM prompt (ablation) | HISE-KT (Qwen-Plus) · Assistment09 | AUC: 0.7734 -> 0.8703 | Improved |
| retrieve similar students by rule-based collaborative filtering instead of random candidates (ablation) | HISE-KT (Qwen-Plus) · Slepemapy | AUC: 0.6937 -> 0.9749 | Improved |
| retrieve similar students by rule-based collaborative filtering instead of random candidates (ablation) | HISE-KT (Qwen-Plus) · Frcsub | AUC: 0.8692 -> 0.9482 | Improved |
| add LLM-based meta-path scorer instead of random-K meta-paths (ablation) | HISE-KT (Qwen-Plus) · Assistment09 | AUC: 0.8593 -> 0.8703 | Improved |
| add LLM-based meta-path scorer instead of random-K meta-paths (ablation) | HISE-KT (Qwen-Plus) · Slepemapy | AUC: 0.9709 -> 0.9749 | Improved |
| use Top-K scoring meta-paths instead of Lowest-K scoring paths (ablation) | HISE-KT (Qwen-Plus) · Statics2011 | AUC: 0.8516 -> 0.8888 | Improved |
| use Lowest-K scoring paths instead of Random-K paths (ablation) | HISE-KT (Qwen-Plus) · Statics2011 | AUC: 0.8805 -> 0.8516 | Hurt |
| add IRT-2PL ability, difficulty and discrimination parameters to the prompt (ablation) | HISE-KT (Qwen-Plus) · Statics2011 | AUC: 0.8447 -> 0.8888 | Improved |
| add IRT-2PL parameters to the prompt (ablation) | HISE-KT (Qwen-Plus) · Slepemapy | AUC: 0.9668 -> 0.9749 | Improved |
| use Qwen-Plus instead of DeepSeek-V3 as the prediction LLM | HISE-KT · Slepemapy | AUC: 0.9383 -> 0.9749 | Improved |
| increase Top-K meta-path instances beyond a moderate value (ablation) | HISE-KT · Parameter sensitivity | AUC | Mixed |
Closest papers by meaning
- Uncertainty-aware Knowledge Tracing (AAAI 2025 · similarity 0.86)
- K12-KGraph: A Curriculum-Aligned Knowledge Graph for Benchmarking and Training Educational LLMs (NeurIPS 2026 · similarity 0.85)
- Predictive, scalable and interpretable knowledge tracing on structured domains (ICLR 2024 · similarity 0.85)
- KScope: A Framework for Characterizing the Knowledge Status of Language Models (NeurIPS 2025 · similarity 0.85)
- CK12: A Rounded K12 Knowledge Graph Based Benchmark for Chinese Holistic Cognition Evaluation (AAAI 2024 · similarity 0.85)
- What Is a Good Question? Assessing Question Quality via Meta-Fact Checking (AAAI 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
