Reflection-Driven Educational Knowledge Graph Expansion Method

DOI

10.2991/978-94-6239-691-3_60How to use a DOI?

Keywords

Domain Knowledge Graph; Large Language Model (LLM); Reflection-Driven Generation; Multidimensional Quality Assessment; Robustness

Abstract

Construction of domain knowledge graphs is critical for intelligent learning and cognitive assessment, yet real-world expansion is often hindered by multi-source noise, leading to semantic drift and structural inconsistency. To address this, we propose RoKG-Agent, a reflection-driven framework for noisy term repair and sibling concept expansion. The method normalizes perturbed terms as semantic anchors and employs multi-dimensional evaluation with iterative optimization to ensure generation quality. We further introduce RoKG-Bench, a hierarchical noise benchmark derived from real-world teaching data. Experiments across multiple foundation models demonstrate that RoKG-Agent significantly improves robustness, structural accuracy, and semantic consistency under severe noise conditions.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Xin Zhang
AU  - Yu Bai
AU  - Guiping Zhang
PY  - 2026
DA  - 2026/05/31
TI  - Reflection-Driven Educational Knowledge Graph Expansion Method
BT  - Proceedings of the 2026 5th International Conference on Educational Innovation and Multimedia Technology (EIMT 2026)
PB  - Atlantis Press
SP  - 605
EP  - 610
SN  - 2667-128X
UR  - https://doi.org/10.2991/978-94-6239-691-3_60
DO  - 10.2991/978-94-6239-691-3_60
ID  - Zhang2026
ER  -