cs.LGNov 17, 2026

Confidence-Based Materials Data Quality Assessment via Domain Knowledge and Graph-Structured Learning

Authors: Sihan Liaing, Ji Zhang, Kehan Wang, Zihao Gao, Wei Ma, Yangang Wang, Zongguo Wang

Organizations: Computer Network Information Center, Chinese Academy of Sciences, Beijing, 100083, China · University of Chinese Academy of Sciences, Beijing, 101408, China · School of Materials and New Energy, Ningxia University, Yinchuan, 750021, China

Abstract

The rapid development of materials machine learning relies on large-scale, heterogeneous datasets, yet model reliability is often compromised by imperfect data from diverse experiments, simulations, and database integration. Real-world materials databases may contain measurement errors, failed calculations, missing or duplicate records, and hidden violations of physicochemical constraints. Such problems are difficult to identify because corrupted samples may remain statistically plausible rather than appearing as explicit outliers. In this paper, we propose a confidence-based materials data quality assessment framework that integrates domain knowledge with graph-structured learning. Data quality is formulated as a continuous confidence estimation problem rather than a binary anomaly identification task, enabling sample-level assessment and historical-reference-based batch comparison. A nine-dimensional feature-quality prior and duplicate-aware sample initialization are calibrated using materials domain rules and statistical evidence. Confidence is further refined through sample-feature bipartite graph propagation and weakly supervised graph attention, while a Memory-GAT module provides historical reference scoring. Experiments on 23 ODDS benchmark datasets show that the generic module achieves an average AUC-ROC of 0.866. In five repeated mixed-contamination experiments on mp_20, rule calibration improves HistGradientBoostingRegressor test R2R^2 from 0.538±0.0080.538\pm0.008 to 0.616±0.0140.616\pm0.014. Graph refinement increases average precision from 0.943±0.0040.943\pm0.004 to 0.953\pm0.004,whiletherule−calibratedconfigurationhasthehighestmeantest, while the rule-calibrated configuration has the highest mean test R^2amongtheevaluatedscreeningconfigurations.FiverepeatedMACEsubsetexperimentsfurthershowanincreaseinamong the evaluated screening configurations. Five repeated MACE subset experiments further show an increase inR^2fromfrom0.590\pm0.053toto0.880\pm0.027$ after rule-based screening.

Figures & tables

Appendix figures & tables3 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Learning Materials Properties from Scarce Labels and Unlabeled Crystals

    Aug 31, 2026Wentao Li, Yizhe Chen, Jiangjie Qiu +3Materials Property PredictionPseudo-Labeling

  2. OmniMatBench: A Human-Calibrated Multimodal Reasoning Benchmark Across 19 Materials Science Subfields

    May 28, 2026Wanhao Liu, Jiaqing Xie, Qian Tan +10Benchmark DesignScientific QA

  3. Physics as the label for measuring and correcting materials reasoning in multimodal models

    Sep 10, 2026Hasan Kurban, Rasul Khanbayov, Mustafa KurbanMultimodal ReasoningMultimodal Model Evaluation