MULTIMODAL PREDICTIVE ANALYTICS: A SYSTEMATIC INTEGRATION OF STRUCTURED DATA, NATURAL LANGUAGE, AND VISUAL MODALITIES FOR ENHANCED DECISION SUPPORT
International Journal of Grid Computing & Applications · 28.03.2026 · 17 (1) · 137-148
DOI: 10.5121/ijgca.2026.17113 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Grid Computing & Applications, online erschienen am 28.03.2026. Verfasst von Sungho Kim und 8 weiteren Autor:innen.
Abstract (Originalsprache)
Contemporary decision-making systems face growing complexity from heterogeneous data streams that span structured tabular records, free-form natural language, and high-dimensional visual content. Unimodal predictive models exploit only a fraction of available information, yielding suboptimal inference. This paper presents a systematic literature review of multimodal predictive analytics, examining 87 peer-reviewed studies published between 2018 and 2025. We analyze fusion architectures (early, late, and hybrid), benchmark reported performance across five application domains—healthcare, finance, sentiment analysis, survival prediction, and autonomous systems—and quantify the accuracy gains attributable to each additional modality. Our synthesis demonstrates that fully multimodal pipelines integrating structured data, natural language processing (NLP), and convolutional or transformer-based vision encoders consistently outperform unimodal baselines by 15.3–19.1 percentage points in classification accuracy and by 0.169 in concordance index for survival tasks. We further identify open challenges in modality alignment, missingdata robustness, and explainability, and outline a forward-looking research agenda to advance trustworthy multimodal decision support.
Zitieren
APA
Kim, S., Rahman, M. M., Rafi, M. A., Shahnawaj, M., Biswas, D., Mridul, A. A., Uddin, H. H., Sikder, M. S., & Mawa, J. (2026). MULTIMODAL PREDICTIVE ANALYTICS: A SYSTEMATIC INTEGRATION OF STRUCTURED DATA, NATURAL LANGUAGE, AND VISUAL MODALITIES FOR ENHANCED DECISION SUPPORT. International Journal of Grid Computing & Applications, 17(1), 137-148. https://doi.org/10.5121/ijgca.2026.17113
Vancouver
Kim S, Rahman MM, Rafi MA, Shahnawaj M, Biswas D, Mridul AA, et al. MULTIMODAL PREDICTIVE ANALYTICS: A SYSTEMATIC INTEGRATION OF STRUCTURED DATA, NATURAL LANGUAGE, AND VISUAL MODALITIES FOR ENHANCED DECISION SUPPORT. International Journal of Grid Computing & Applications. 2026;17(1):137-148. doi:10.5121/ijgca.2026.17113
BibTeX
@article{Kim2026,
author = {Kim, Sungho and Rahman, Mohammad Mahmudur and Rafi, Mainuddin Adel and Shahnawaj, Md and Biswas, Debabrata and Mridul, Abdul Ahad and Uddin, Himel Hemayet and Sikder, Mohammad Somon and Mawa, Jannatul},
title = {MULTIMODAL PREDICTIVE ANALYTICS: A SYSTEMATIC INTEGRATION OF STRUCTURED DATA, NATURAL LANGUAGE, AND VISUAL MODALITIES FOR ENHANCED DECISION SUPPORT},
journal = {International Journal of Grid Computing & Applications},
year = {2026},
volume = {17},
number = {1},
pages = {137--148},
doi = {10.5121/ijgca.2026.17113}
}
EUVIMED