What question did this study set out to answer?

The aim is to enhance social media popularity prediction for short-form video content using a multi-modal approach.

April 26, 2026Applied SciencesOpen Access

FUSEPOP: A Multi-Modal Fusion with Mutual Information Weighting and Stacked Ensemble for Social Media Popularity Prediction

View Full Paper

Ask AI

Bookmark

Authors

Randomized trial demonstrates improved social media popularity prediction using a novel multi-modal framework, indicating effective data integration.

The aim is to enhance social media popularity prediction for short-form video content using a multi-modal approach.
Developed FUSEPOP architecture utilizing ResNet-50 for visual features and XLM-RoBERTa for textual information.
Employed mutual information weighting and stacked generalization with logistic regression as a meta-learner, trained via stratified 5-fold cross-validation.
Integrated predictions from random forest, XGBoost, and neural network classifiers.
Achieved an accuracy of 0.980 and an average F1-score of 0.964 on selected features.
Outperformed benchmark models significantly on the primary dataset.
Remained competitive on a literature-aligned alternative configuration.

ŞENCAN et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

Randomized trial demonstrates improved social media popularity prediction using a novel multi-modal framework, indicating effective data integration.