What question did this study set out to answer?

The aim is to develop a multimodal framework that enhances URL classification accuracy by integrating visual, structural, and semantic representations.

July 5, 2026Scientific ReportsOpen Access

Multimodal fusion for malicious URL detection using visual, structural, and semantic representations

View Full Paper

Ask AI

Bookmark

Authors

JLJ H LeeKumoh National Institute of Technology HKHyun KwonKorea Military Academy

Discussion

Member takes

Implication

Randomized trial demonstrates enhanced URL classification accuracy in various representations, indicating a significant improvement over traditional methods.

Key Points

The aim is to develop a multimodal framework that enhances URL classification accuracy by integrating visual, structural, and semantic representations.
URLs converted into grayscale images for CNN processing based on ResNet architecture.
Structural features analyzed using tree-based classifiers XGBoost and CatBoost.
Semantic insights obtained through a fine-tuned DistilBERT model, integrated using a deep neural network for final classification.
Achieved an F1-score of 0.9189 and AUC of 0.9805, indicating significant performance improvement over existing methods.
The multimodal framework outperformed traditional single-modal classifications, enhancing accuracy.

Cite This Study

Lee et al. (2026) studied this question.

synapsesocial.com/papers/6a49f3b0f5d1d45b287ffae0 https://doi.org/10.1038/s41598-025-24220-3

View Full Paper

Ask AI

Bookmark

Implication

Randomized trial demonstrates enhanced URL classification accuracy in various representations, indicating a significant improvement over traditional methods.

Multimodal fusion for malicious URL detection using visual, structural, and semantic representations

Discussion

Member takes

Implication

Key Points

Cite This Study

Multimodal fusion for malicious URL detection using visual, structural, and semantic representations

Discussion

Member takes

Implication

Key Points

Cite This Study

Also Consider

Also Consider