EUVIMEDEUVIMED entdecken
Crossref

Research on a Closed-Loop Multi-Agent Collaborative Decision-Making and Personalized Knowledge Generation System for Vertical Domains

Yonghui Zheng, Ruiheng Zhang, Chao Geng, Xiaojun Fan, Hanxiao Chang, Mingming Gong

Journal of Big Data and Computing · 2026-09 · 4 (3) · 183-192

DOI: 10.62517/jbdc.202601323 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Big Data and Computing, 2026-09. Verfasst von Yonghui Zheng und 5 weiteren Autor:innen. Die Arbeit nennt 16 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Traditional online learning systems still have several limitations. These limitations include inflexible learning paths, coarse-grained resource recommendations, delayed feedback, and limited intelligent interaction. This study develops a personalized learning system. The system combines multi-agent collaboration with dynamic learner profiles. The system first identifies learner requests through an intent recognition module. The intent recognition result determines which agent should handle the task. Different agents are responsible for different learning activities. These agents include the Question-Answering Agent, Path Planning Agent, Resource Recommendation Agent, Question Bank Generation Agent, Document Parsing Agent, Code Tutoring Agent, and Learning Evaluation Agent. The system updates the learner profile according to learning behaviors, knowledge mastery states, learning goals, and knowledge gaps. The learner profile is combined with the knowledge graph. This combination supports staged path planning and resource recommendation. The system uses error-feedback mechanisms, Feynman learning logs, and event-triggered mechanisms to build a closed learning loop. The loop connects diagnosis, path generation, resource recommendation, and adaptive feedback adjustment. Experiments and case studies show that the system can generate more reasonable learning paths and provide better resource recommendations. The proposed system provides a practical reference for personalized intelligent learning systems.

Zitieren

APA

Zheng, Y., Zhang, R., Geng, C., Fan, X., Chang, H., & Gong, M. (2026). Research on a Closed-Loop Multi-Agent Collaborative Decision-Making and Personalized Knowledge Generation System for Vertical Domains. Journal of Big Data and Computing, 4(3), 183-192. https://doi.org/10.62517/jbdc.202601323

Vancouver

Zheng Y, Zhang R, Geng C, Fan X, Chang H, Gong M. Research on a Closed-Loop Multi-Agent Collaborative Decision-Making and Personalized Knowledge Generation System for Vertical Domains. Journal of Big Data and Computing. 2026;4(3):183-192. doi:10.62517/jbdc.202601323

BibTeX
@article{Zheng2026,
  author = {Zheng, Yonghui and Zhang, Ruiheng and Geng, Chao and Fan, Xiaojun and Chang, Hanxiao and Gong, Mingming},
  title = {Research on a Closed-Loop Multi-Agent Collaborative Decision-Making and Personalized Knowledge Generation System for Vertical Domains},
  journal = {Journal of Big Data and Computing},
  year = {2026},
  volume = {4},
  number = {3},
  pages = {183--192},
  doi = {10.62517/jbdc.202601323}
}
Kostenlos starten →