Explainable Artificial Intelligence in Computational Materials Science: Mapping the Landscape of CALPHAD-Based Design

The integration of explainable artificial intelligence (XAI) with computational materials science has emerged as a transformative paradigm for accelerating materials discovery and understanding structure– property relationships. This investigation executes an exhaustive bibliometric evaluation to map the scholarly convergence of computational materials science and explainable artificial intelligence (XAI). Using a systematic search in the Web of Science Core Collection, 157 relevant articles published between 2019 and 2026 were identified. VOSviewer software was employed to map co-authorship networks (authors, organizations, countries), keyword co-occurrence, and citation networks (documents, sources, authors). The results reveal a rapidly growing field, with annual publications rising from 2 in 2019 to 34 in 2026. Co-authorship analysis shows a dominant cluster of researchers centered around CALPHAD-based alloy design, and a distinct international collaboration network led by the USA, Germany, and South Korea. Keyword co-occurrence highlights four thematic clusters: (i) explainable AI and interpretable machine learning, (ii) material informatics and design, (iii) SHAP and feature selection, and (iv) specific techniques such as deep learning and friction stir welding. Citation analysis indicates that the most influential works focus on high-strength aluminum alloy design and CALPHAD-driven inverse predictions. The sparse citation linkages among documents and sources suggest a fragmented early-stage field. This bibliometric overview provides a foundation for researchers to identify emerging trends, influential contributors, and thematic gaps, thereby guiding future interdisciplinary efforts in explainable computational materials science.