PPN – Pedagogy Pathfinder Network: A Multi-Agent, Retrieval-Augmented AI Framework for Automated Academic Administration Plan Generation

This survey examined 17 studies in four areas: retrieval-augmented generation, LLM-based multi-agent orchestration, Bloom’s-taxonomy and CO–PO automation, and AI-assisted lesson planning and assessment. The components on which a system for generating NBA-oriented Academic Administration Plans would depend, namely retrieval grounding, staged agent collaboration, outcome classification and mapping, and lesson-plan and rubric generation, have each been studied, but under different conditions and with different evaluation methods. Within the reviewed set, outcome-mapping work concentrates on text that already exists, and none of the studies combines document-grounded generation, agent orchestration and automated compliance checking for Indian accreditation documents. Four gap areas follow: integration and compliance validation, small single-institution datasets, error propagation and scalability, and the unmeasured cost of human review. Addressing them calls for shared benchmarks, tests of layered agent communication, and comparison of generated plans with manually prepared ones. Integrating RAG, multi-agent orchestration and compliance validation may reduce the documentation burden on faculty, but whether it does so reliably, and at what review cost, has still to be shown by independent evaluation.

[1]     P. Lewis, E. Perez, A. Piktus, F. Petroni, V. Karpukhin, N. Goyal, H. Küttler, M. Lewis, W. Yih, T. Rocktäschel, S. Riedel, and D. Kiela, “Retrieval-augmented generation for knowledge-intensive NLP tasks,” in Advances in Neural Information Processing Systems (NeurIPS), vol. 33, 2020, pp. 9459–9474. [2]     Y. Gao, Y. Xiong, X. Gao, K. Jia, J. Pan, Y. Bi, Y. Dai, J. Sun, and H. Wang, “Retrieval-augmented generation for large language models: A survey,” arXiv:2312.10997, 2023 (preprint). [3]     P. Zhao, H. Zhang, Q. Yu, Z. Wang, Y. Geng, F. Fu, L. Yang,W. Zhang, J. Jiang, and B. Cui, “Retrieval-augmented generation for AI-generated content: A survey,” arXiv:2402.19473, 2024 (preprint). [4]     W. Fan, Y. Ding, L. Ning, S. Wang, H. Li, D. Yin, T.-S. Chua, and Q. Li, “A survey on RAG meeting LLMs: Towards retrieval-augmented large language models,” in Proc. 30th ACM SIGKDD Conf. Knowledge Discovery and Data Mining (KDD), 2024, pp. 6491–6501. [5]     Q. Wu, G. Bansal, J. Zhang, Y. Wu, S. Zhang, E. Zhu, B. Li,L. Jiang, X. Zhang, and C. Wang, “AutoGen: Enabling next-gen LLM applications via multi-agent conversation framework,” arXiv:2308.08155, 2023 (preprint). [6]     S. Hong, X. Zheng, J. Chen, Y. Cheng, J. Wang, C. Zhang, Z. Wang, S. K. S. Yau, Z. Lin, L. Zhou, C. Ran, L. Xiao, and C. Wu, “MetaGPT: Meta programming for a multi-agent collaborative framework,” arXiv:2308.00352, 2023 (preprint). [7]     C. Qian, W. Liu, H. Liu, N. Chen, Y. Dang, J. Li, C. Yang,W. Chen, Y. Su, X. Cong, J. Xu, D. Li, Z. Liu, and M. Sun, “ChatDev: Communicative agents for software development,” in Proc. 62nd Annu. Meeting Assoc. Computational Linguistics (ACL), vol. 1, 2024, pp. 15174–15186. [8]     T. Guo et al., “Large language model based multi-agents: A survey of progress and challenges,” in Proc. 33rd Int. Joint Conf. Artificial Intelligence (IJCAI), 2024, pp. 8048–8057. [9]     S. Elkins, E. Kochmar, J. C. K. Cheung, and I. Serban, “How teachers can use large language models and Bloom’s taxonomy to create educational quizzes,” in Proc. AAAI Conf. Artificial Intelligence, vol. 38, no. 21, 2024, pp. 23378–23386. [10]   A. Waheed, M. Goyal, N. Mittal, D. Gupta, A. Khanna, andM. Sharma, “BloomNet: A robust transformer based model for Bloom’s learning outcome classification,” in Proc. 4th Int. Conf. Natural Language and Speech Processing (ICNLSP), 2021, pp. 209–218.