Identification of Multilingual Offense and Troll from Social Media Memes Using Weighted Ensemble of Multimodal Features

Keyphrases

• Social Media 100%
• Multimodal Features 100%
• Memes 100%
• Weighted Ensemble 100%
• Multimodal Model 50%
• F1 Score 33%
• Ensemble Methods 33%
• Offensive 16%
• Comparative Analysis 16%
• Benchmark Dataset 16%
• Human Emotions 16%
• Evade 16%
• Detrimental Effects 16%
• Visual Model 16%
• ResNet50 16%
• Joint Modeling 16%
• Offensive Meme 16%
• Fusion Approach 16%
• XLM-R 16%
• Text-to-image 16%
• DistilBERT 16%
• Multilingual BERT 16%
• VGG19 16%
• VGG16 16%
• Early-late 16%
• Visual Features 16%
• Unimodal Model 16%
• Ensemble Model 16%
• R-model 16%
• Social Harmony 16%
• Early Fusion 16%
• Content Polluters 16%
• Text Features 16%
• Late Fusion 16%

Computer Science

• Ensemble Technique 100%
• Bidirectional Encoder Representations From Transformers 50%
• Utmost Importance 50%
• Social Harmony 50%
• Textual Feature 50%
• Visual Feature 50%
• VGG-19 Convolutional Neural Network 50%
• ResNet50 50%
• DistilBERT 50%
• VGG16 50%