Document chunking fundamentally impacts Retrieval-Augmented Generation (RAG) by determining how source materials are segmented before indexing.Despite evidence that Large Language Models (LLMs) are sensitive to the layout and structure of retrieved data, there is currently no framework to analyze the impact of different chunking methods.In this paper, we introduce a novel methodology that defines essential characteristics of the chunking process at three levels: intrinsic passage properties, extrinsic passage properties, and passages-document coherence.We propose HOPE (Holistic Passage Evaluation), a domain-agnostic, automatic evaluation metric that quantifies and aggregates these characteristics.Our empirical evaluations across seven domains demonstrate that the HOPE metric correlates significantly (𝜌 > 0.13) with various RAG performance indicators, revealing contrasts between the importance of extrinsic and intrinsic properties of passages.Semantic independence between passages proves essential for system performance with a performance gain of up to 56.2% in factual correctness and 21.1% in answer correctness.On the contrary, traditional assumptions about maintaining concept unity within passages show minimal impact.These findings provide actionable insights for optimizing chunking strategies, thus improving RAG system design to produce more factually correct responses.

The work presented here is a high dimensional color image encryption architecture (HDIEA) founded on the Lorenz-Gauss-Logistic (LGL) encryption algorithm.The primary objective is to demonstrate that both the proposed novel five-dimensional (5D) Gauss-Logistic and four-dimensional (4D) Lorenz system are operating in a hyperchaotic condition.The visual study of their most important characteristics, such as the sensitivity of the starting value of both maps and the Lyapunov exponent of the 5D Gauss Logistic map, is carried out.The Runge-Kutta technique is used to discretise the Lorenz system in order to construct a pseudo-random sequence generation for the control parameter that has a greater degree of randomness.The 5D Gauss-Logistic system is then selected to serve as the principal hyper-chaotic mapping scheme.The simulation results demonstrate that the suggested image encryption method is successful according to the NIST test and has powerful anti-attack, a larger key space as large as 2 847 , which is prone to multiple attacks, and key sensitivity capabilities.Also, the pixel correlation reached -0.0019, -0.0016, and -0.0069, while the information entropy was at 7.9996.This demonstrates the excellent scrambling effect of the proposed approach, which is capable of greatly improving the color image security performance.