In this work, we have developed a novel computational approach for the investigation of structure-process-property relationships in materials through use of a Network Component Analysis (NCA) algorithm when knowledge of the underlying structural topology is only partially known. This technique can potentially provide a robust understanding of how processing conditions affect structural domains across multiple length scales.  For instance, current modeling techniques such as molecular mechanics or continuum-based approximations are limited in their ability to model materials due to their computationally intensive nature, as well as the need for simplifying assumptions when coarse-graining across multiple scales. Our approach can be driven by either simulation or experimental data, that is organized into a (M-by-N) matrix E composed of N data points obtained from M distinctly prepared material samples.  E is then decomposed into a topology matrix A (M x L), and perturbation matrix P (L x N) by the construct E=AP.  Identification of the L parameters that characterize the system, as well as their physical meaning, describe this topology.  This technique has been successfully applied to analysis of DNA microarray data for connecting gene expression and transcription factor activity.  Our goal is to establish similar fundamental connectivity relationships between material structure and processing conditions.