An AI-Driven Intelligent Decision-Making Framework for Digital Media Analytics, Automated Content Intelligence, Audience Prediction, and Data-Driven Communication Strategies
Computer and Decision Making: An International Journal · 04.11.2026 · 3 (6S) · 3552-3563
DOI: 10.1922/comdem.v3i6s.19098 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computer and Decision Making: An International Journal, online erschienen am 04.11.2026. Verfasst von Dr Swapnadip De und 5 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Based on the above, this study models and tests the intelligent decision-making process using AI by analyzing the digital media, creating the automated content intelligence, forecasting the audience's reactions, using explainable artificial intelligence, and providing communication support based on data. The analysis was done on a dataset of 19,885 social media engagements publicly available from Instagram, Facebook, Twitter/X and LinkedIn. This framework consisted of descriptive analytics, linguistic feature engineering, supervised machine learning, incremental-value testing, SHAP based explainability. The ridge regression, the random forest, the XGBoost algorithm, and the CatBoost algorithm were tested using a chronological 80:20 split of the data for training and testing, respectively, and the performance of these models was assessed using the mean absolute error, the root mean squared error, and the coefficient of determination. Average engagement was similar across platforms and post types, with only minor variations observed. Modelling predictive performance was poor with all models having R² values near zero or negative. The content-intelligence Ridge model had the lowest RMSE value of 13.920 and the lowest MAE value of 10.201 for contextual CatBoost. The addition of automated content-intelligence features did not prove to be beneficial in predicting beyond the context information. Audience Interests, Month, Audience Age, Post Hour, and Audience Location were all deemed important model features by SHAP; however, these interpretations were done with some caution due to the limited predictive validity. The study thus lends support to a decision-support architecture that is validation gated, in which the generation of communication recommendations is contingent on predictive evidence being above or below a predetermined baseline.
Zitieren
APA
De, D. S., S, N. G., Toshtemirovna, R. G., Prasad, M. T. S. D., Acharjee, D. P. B., & Abazadze, G. (2026). An AI-Driven Intelligent Decision-Making Framework for Digital Media Analytics, Automated Content Intelligence, Audience Prediction, and Data-Driven Communication Strategies. Computer and Decision Making: An International Journal, 3(6S), 3552-3563. https://doi.org/10.1922/comdem.v3i6s.19098
Vancouver
De DS, S NG, Toshtemirovna RG, Prasad MTSD, Acharjee DPB, Abazadze G. An AI-Driven Intelligent Decision-Making Framework for Digital Media Analytics, Automated Content Intelligence, Audience Prediction, and Data-Driven Communication Strategies. Computer and Decision Making: An International Journal. 2026;3(6S):3552-3563. doi:10.1922/comdem.v3i6s.19098
BibTeX
@article{De2026,
author = {De, Dr Swapnadip and S, Nanda Gopal and Toshtemirovna, Rabbimova Gulnara and Prasad, Mr Tripurari Sai Durga and Acharjee, Dr Purnendu Bikash and Abazadze, Giorgi},
title = {An AI-Driven Intelligent Decision-Making Framework for Digital Media Analytics, Automated Content Intelligence, Audience Prediction, and Data-Driven Communication Strategies},
journal = {Computer and Decision Making: An International Journal},
year = {2026},
volume = {3},
number = {6S},
pages = {3552--3563},
doi = {10.1922/comdem.v3i6s.19098}
}
EUVIMED