NCERT-tailored RAG preprocessing and figure-aware indexing improve retrieval precision over a generic-chunking baseline, per the authors' assertion¶

Claim · All claims
Evidence · 1 study · design r1 · q1
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person · recheck by 2029: tool and model accuracy changes quickly

Subclaims¶

q1 i? The authors assert that targeted text cleaning plus figure-aware document extraction improves retrieval precision and contextual grounding compared to a baseline RAG pipeline using generic text chunking and embedding. → Siddhanth Sridhar 2026

Evidence¶

Siddhanth Sridhar 2026¶

Siddhanth Sridhar, Shreya Chaurasia, Baddela Sai Yaswantha Reddy, Deepak Parmar, and Dr. Shylaja S S. (2026). Dynamic Learning Solutions: A System for Personalized Educational Video Generation. arXiv preprint. https://arxiv.org/abs/2609.14408 doi:10.48550/arxiv.2609.14408

q1 · i? · design · r1

The authors' design assertion in the methodology section describing their NCERT-optimized RAG pipeline; no retrieval evaluation, metric, or test is printed. The claimed gain is "retrieval precision and contextual grounding" versus a generic-chunking baseline.

"This optimization improves retrieval precision and contextual grounding for educational queries compared to a baseline RAG pipeline that only relies on generic text chunking and embedding."

Discussion¶

Related Claims¶

• Retrieval context has a sweet spot at Top-K = 10: raising it to 15 did not improve recall but significantly worsened context precision via the distraction issue — related
• Optimal RAG configuration differs by discipline: STEM subjects perform best with 512-character chunks and temperature 0.3 (faithfulness 0.97), while humanities require 1000-character chunks and temperature 0.1 (context recall 1.00, faithfulness 0.98) — related
• Pipeline components play complementary roles: hierarchical generation has the largest effect, grounding improves specificity and factual accuracy, and cluster-informed TOC planning improves structure and audience fit — related
• Indexing LLM-generated artifacts nearly doubles retrieval recall on analytical queries compared with transcript-only retrieval, while direct queries perform comparably across configurations — related