Corporate Performance Evaluation Using Management Accounting Information Systems

Abstract

This research introduces a novel, cross-disciplinary framework for corporate performance evaluation that synthesizes principles from management accounting information systems (MAIS), computational linguistics, and network theory. Departing
from traditional financial ratio analysis, we propose the Integrated Performance
Resonance (IPR) model, which treats an organization as a dynamic information
ecosystem. The model captures not only quantitative financial flows but also the
qualitative, narrative data embedded in managerial reports, internal communications, and strategic documents. Our methodology employs a hybrid technique
combining sentiment-aware natural language processing (NLP) to extract latent
strategic intent and operational sentiment from unstructured text, with temporal
graph networks to model the flow and influence of accounting information across
organizational units. We apply this framework to a unique longitudinal dataset
from a multinational corporation, analyzing over 15,000 internal documents alongside traditional accounting data over a five-year period. The results demonstrate
that the IPR model identifies performance drivers and risk precursors with 37