Integrating Explainable AI with Advanced Data Visualization: Enhancing Transparency and Decision-Making in Complex Systems

Abstract

[1]. Arh, T., & Blaži, B. J. (2007). Application of Multi-Attribute Decision Making Approach to Learning Management Systems Evaluation . JOURNAL OF COMPUTERS, 2(10), 28-37. [2]. BÂRA, A., BOTHA, I., LUNGU, I., & OPREA, S. V. (2013). Decision Support System in National Power Companies A Practical Example (Part I) . Database Systems Journal, IV(1), 37-45. The rapid adoption of artificial intelligence in high-stakes domains such as healthcare, finance, and cybersecurity has intensified concerns about the opacity of complex models, which undermines stakeholder trust and slows adoption. Explainable artificial intelligence (XAI) addresses this by producing justifications for model decisions, yet such justifications are typically delivered as raw textual or numerical output that many stakeholders find difficult to interpret. This paper proposes a conceptual framework that integrates XAI with advanced data visualization, treating visualization as a first-class layer of the explanation pipeline rather than an afterthought: post-hoc attribution methods supply explanation content, and purpose-designed visual encodings — heatmaps, feature-importance dashboards, and decision-path graphs — supply a cognitively efficient form. Beyond the framework, the paper specifies a standardized, falsifiable evaluation protocol spanning three benchmark domains and seven measurable constructs — explanation satisfaction, objective understanding, decision accuracy, cognitive load, subjective trust, calibrated trust, and decision confidence — with directional hypotheses grounded in perceptual and cognitive-explanation theory. A structured results-reporting design is included to support immediate empirical application. The primary contribution is the framework, the measurement standard, and the evaluation protocol.