{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,17]],"date-time":"2026-06-17T04:51:46Z","timestamp":1781671906790,"version":"3.54.5"},"reference-count":33,"publisher":"MDPI AG","issue":"6","license":[{"start":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T00:00:00Z","timestamp":1781568000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Computers"],"abstract":"<jats:p>Affected by inter-annotator cognitive differences, fatigue effects, and data poisoning, training data inevitably contains a certain proportion of noise, which severely impairs model performance. Traditional manual verification is costly and inefficient, while existing automatic detection methods generally suffer from limited precision, poor interpretability, and insufficient robustness. This paper proposes a noise label detection and correction method based on Bayesian weighted consensus inference. First, an ensemble of multiple lightweight heterogeneous models is constructed, and model prior knowledge and dataset noise are obtained on a clean validation set. Second, the model ensemble predicts noisy samples to extract two-dimensional consensus evidence. Then, prior knowledge and consensus evidence are fused, and the posterior probability of label noise is calculated via Bayesian inference to generate correction suggestions. Finally, high-confidence noisy labels are precisely screened based on the posterior probability threshold. Experimental results on three datasets show that the proposed method achieves a precision of 96.50%, a recall of 98.61%, an F1-score of 97.54%, and a correction accuracy of 95.53%, with improvements of 5\u201320% over mainstream methods. With a computational cost comparable to that of basic ensemble methods, the proposed approach achieves a favorable balance among precision, robustness, and interpretability. It thus offers a promising and cost-effective solution for automated quality control of large-scale annotated datasets, especially in text classification tasks.<\/jats:p>","DOI":"10.3390\/computers15060387","type":"journal-article","created":{"date-parts":[[2026,6,17]],"date-time":"2026-06-17T03:27:21Z","timestamp":1781666841000},"page":"387","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Noise Label Detection and Correction via Bayesian Weighted Consensus Inference"],"prefix":"10.3390","volume":"15","author":[{"given":"Qi","family":"Yang","sequence":"first","affiliation":[{"name":"Naval University of Engineering, Wuhan 430033, China"},{"name":"Navy Unit 91646, Jiujiang 320005, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jing","family":"Li","sequence":"additional","affiliation":[{"name":"Naval University of Engineering, Wuhan 430033, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Aoyun","family":"Zhu","sequence":"additional","affiliation":[{"name":"Naval University of Engineering, Wuhan 430033, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hao","family":"Chen","sequence":"additional","affiliation":[{"name":"Naval University of Engineering, Wuhan 430033, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,6,16]]},"reference":[{"key":"ref_1","first-page":"1","article-title":"Applications and future prospects of computer vision in aquaculture","volume":"53","author":"Peng","year":"2026","journal-title":"Fish. Mod."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"2488120","DOI":"10.1080\/21642583.2025.2488120","article-title":"A survey on learning from data with label noise via deep neural networks","volume":"13","author":"Song","year":"2025","journal-title":"Syst. Sci. Control Eng."},{"key":"ref_3","doi-asserted-by":"crossref","first-page":"310","DOI":"10.1007\/s10462-025-11293-9","article-title":"Benchmarking machine learning methods for the identification of mislabeled data","volume":"58","author":"Nazaretyan","year":"2025","journal-title":"Artif. Intell. Rev."},{"key":"ref_4","unstructured":"Saeeda, H., Johansson, T., Mohamad, M., and Knauss, E. (2025). Data annotation quality problems in AI-enabled perception system development. arXiv."},{"key":"ref_5","first-page":"1","article-title":"Exploration and prospects of large language models in traditional chinese medicine orthopedics and traumatology","volume":"39","author":"Feng","year":"2026","journal-title":"China J. Orthop. Traumatol."},{"key":"ref_6","first-page":"17","article-title":"The key issues and evaluation methods for constructing agricultural pest and disease image datasets: A review","volume":"5","author":"Guan","year":"2023","journal-title":"Smart Agric."},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"G\u1f79mez-Zamanillo, L., Portilla, N., Pic\u1f79n, A., Egusquiza, I., Navarra-Mestre, R., Elola, A., and Bereciartua-Perez, A. (2025). Reducing annotation effort in agricultural data: Simple and fast unsupervised coreset selection with DINOv2 and K-means. Front. Plant Sci., 16.","DOI":"10.3389\/fpls.2025.1546756"},{"key":"ref_8","unstructured":"Li, Y.H. (2025). Research and Application of Multimodal Relation Extraction Algorithm, University of Electronic Science and Technology of China."},{"key":"ref_9","doi-asserted-by":"crossref","first-page":"3629","DOI":"10.11834\/jig.220953","article-title":"A survey on adversarial training for robust learning","volume":"28","author":"Sui","year":"2023","journal-title":"J. Image Graph."},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"2672","DOI":"10.11834\/jig.230412","article-title":"A review of adversarial examples for optical character recognition","volume":"29","author":"Guo","year":"2024","journal-title":"J. Image Graph."},{"key":"ref_11","first-page":"758","article-title":"Research progress on nuclear accident emergency decision-making based on deep learning","volume":"46","author":"Ma","year":"2026","journal-title":"Nucl. Electron. Detect. Technol."},{"key":"ref_12","unstructured":"Li, H.D. (2024). Research on Enhancement Method of Android Malware Detection Model Based on Uncertainty, Beijing University of Posts and Telecommunications."},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"1373","DOI":"10.1613\/jair.1.12125","article-title":"Confident learning: Estimating uncertainty in dataset labels","volume":"70","author":"Northcutt","year":"2021","journal-title":"J. Artif. Intell. Res."},{"key":"ref_14","first-page":"59","article-title":"Ensemble classification methods for imbalanced data streams","volume":"61","author":"Zhu","year":"2025","journal-title":"Comput. Eng. Appl."},{"key":"ref_15","first-page":"2365","article-title":"Survey on data integration technologies for relational data and knowledge graph","volume":"34","author":"Gao","year":"2023","journal-title":"J. Softw."},{"key":"ref_16","first-page":"1082","article-title":"Ensemble learning methods: A survey","volume":"40","author":"Xu","year":"2018","journal-title":"J. Yunnan Univ. (Nat. Sci. Ed.)"},{"key":"ref_17","first-page":"1","article-title":"Status and challenges of large language models applications in vertical domains","volume":"61","author":"Ji","year":"2025","journal-title":"Comput. Eng. Appl."},{"key":"ref_18","unstructured":"Goodfellow, I., Shlens, J., and Szegedy, C. (2015). Explaining and harnessing adversarial examples. arXiv."},{"key":"ref_19","unstructured":"Zhang, C., Bengio, S., Hardt, M., Recht, B., and Vinyals, O. (2016). Understanding deep learning requires rethinking generalization. arXiv."},{"key":"ref_20","unstructured":"Rolnick, D., Veit, A., Belongie, S., and Shavit, N. (2018). Deep learning is robust to massive label noise. arXiv."},{"key":"ref_21","first-page":"111","article-title":"A survey on the confidence mechanism research in machine learning","volume":"38","author":"Jiang","year":"2014","journal-title":"J. Beijing Jiaotong Univ."},{"key":"ref_22","unstructured":"(2026, June 11). Cleanlab Open-Source Documentation. Available online: https:\/\/docs.cleanlab.ai\/stable\/."},{"key":"ref_23","unstructured":"Guo, C., Pleiss, G., Sun, Y., and Weinberger, K.Q. (2017). On calibration of modern neural networks. Proceedings of the 34th International Conference on Machine Learning, Sydney, Australia, PMLR."},{"key":"ref_24","doi-asserted-by":"crossref","unstructured":"Feng, W., Quan, Y., and Dauphin, G. (2020). Label noise cleaning with an adaptive ensemble method based on noise detection metric. Sensors, 20.","DOI":"10.3390\/s20236718"},{"key":"ref_25","unstructured":"Ren, M., Zeng, W., Yang, B., and Urtasun, R. (2018). Learning to reweight examples for robust deep learning. Proceedings of the 35th International Conference on Machine Learning (ICML\u201918), Stockholm, Sweden, PMLR."},{"key":"ref_26","unstructured":"Li, J., Socher, R., and Hoi, S.C.H. (May, January 26). DivideMix: Learning with Noisy Labels as Semi-Supervised Learning. Proceedings of the International Conference on Learning Representations (ICLR), Addis Ababa, Ethiopia."},{"key":"ref_27","doi-asserted-by":"crossref","unstructured":"Zhao, Y., Lv, W., Xu, S., Wei, J., Wang, G., Dang, Q., Liu, Y., and Chen, J. (2024, January 10\u201317). DETRs Beat YOLOs on Real-time Object Detection. Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA.","DOI":"10.1109\/CVPR52733.2024.01605"},{"key":"ref_28","unstructured":"(2024). Deepseek-v2: A strong, economical, and efficient mixture-of-experts language model. arXiv."},{"key":"ref_29","unstructured":"Bai, J.Z., Bai, S., Chu, Y.F., Cui, Z.Y., Dang, K., Deng, X.D., Fan, Y., Ge, W., Han, Y., and Huang, F. (2023). Qwen technical report. arXiv."},{"key":"ref_30","unstructured":"OpenAI (2023). GPT-4 Technical Report, OpenAI."},{"key":"ref_31","first-page":"1","article-title":"Exploring the limits of transfer learning with a unified text-to-text transformer","volume":"21","author":"Raffel","year":"2020","journal-title":"J. Mach. Learn. Res."},{"key":"ref_32","doi-asserted-by":"crossref","unstructured":"Lewis, M., Liu, Y., Goyal, N., Ghazvininejad, M., Mohamed, A., Levy, O., Stoyanov, V., and Zettlemoyer, L. (2019). BART: Denoising sequence-to-sequence pre-training for natural language generation, translation, and comprehension. arXiv.","DOI":"10.18653\/v1\/2020.acl-main.703"},{"key":"ref_33","unstructured":"Liu, Y.H., Ott, M., Goyal, N., Du, J.F., Joshi, M., Chen, D.Q., Levy, O., Lewis, M., Zettlemoyer, L., and Stoyanov, V. (2019). RoBERTa: A robustly optimized BERT pretraining approach. arXiv."}],"container-title":["Computers"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2073-431X\/15\/6\/387\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,17]],"date-time":"2026-06-17T03:55:45Z","timestamp":1781668545000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2073-431X\/15\/6\/387"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,16]]},"references-count":33,"journal-issue":{"issue":"6","published-online":{"date-parts":[[2026,6]]}},"alternative-id":["computers15060387"],"URL":"https:\/\/doi.org\/10.3390\/computers15060387","relation":{},"ISSN":["2073-431X"],"issn-type":[{"value":"2073-431X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,16]]}}}