Adaptive Tutoring & Heuristic Engagement Network Architecture(ATHENA)
Kuberalakshmi K
, Rameela K , Keerthana G , Subhasri V , Dr. C. Anbuananth
Keywords—Adaptive Learning, AI Tutoring, ChromaDB, Hallucination Elimination, Intelligent Tutoring System, LLM, RAG, Retrieval-Augmented Generation, FastAPI, Groq, Personalized Education.
ATHENA (Adaptive Tutoring & Heuristic Engagement Network Architecture) is an AI-driven intelligent tutoring system that personalizes learning paths and delivers hallucination-free educational responses grounded exclusively in verified academic textbook content. Built on a Retrieval-Augmented Generation (RAG) architecture, ATHENA embeds academic documents using BAAI/bge-small-en-v1.5, stores them in a ChromaDB vector database, and retrieves contextually relevant passages via Maximum Marginal Relevance (MMR) before generating responses through the Groq LLM. Every answer includes exact page-number citations, ensuring full verifiability and academic transparency. The system dynamically adapts quiz difficulty and study recommendations based on real-time student performance tracking. Implemented as a full-stack web application using React, FastAPI, and PostgreSQL, ATHENA achieves an end-to-end query latency of 16.04 ms on average with a RAGAS Faithfulness score of 0.96. ATHENA directly addresses three critical failures of current educational AI: non-adaptive instruction, AI hallucination, and absence of source citations.
"Adaptive Tutoring & Heuristic Engagement Network Architecture(ATHENA)", JETNR - JOURNAL OF EMERGING TRENDS AND NOVEL RESEARCH (www.JETNR.org), ISSN:2984-9276, Vol.4, Issue 4, page no.a484-a488, April-2026, Available :https://rjpn.org/JETNR/papers/JETNR2604059.pdf
Volume 4
Issue 4,
April-2026
Pages : a484-a488
Paper Reg. ID: JETNR_233624
Published Paper Id: JETNR2604059
Downloads: 00047
Research Area: Science and Technology
Country: cuddalore, Tamilnadu, India
ISSN: 2984-9276 | IMPACT FACTOR: 9.87 Calculated By Google Scholar | ESTD YEAR: 2023
An International Scholarly Open Access Journal, Peer-Reviewed, Refereed Journal Impact Factor 9.87 Calculate by Google Scholar and Semantic Scholar | AI-Powered Research Tool, Multidisciplinary, Monthly, Multilanguage Journal Indexing in All Major Database & Metadata, Citation Generator
Publisher: RJPN (IJPublication) Janvi Wave
