A Transformer Based Approach for Abuse Detection in Code Mixed Indic Languages.

Keyphrases

• Indo-Aryan Languages 100%
• Transformer-based 100%
• XLM-RoBERTa 100%
• Cyberspace 100%
• Code-mixing 100%
• Combined Model 100%
• Web User 100%
• Classical Machine Learning Model 100%
• Multilingual Text 100%
• Abuse Detection 100%
• F1 Score 100%
• Social Media Platforms 50%
• Suitable Solutions 50%
• Data Combination 50%
• Language Model 50%
• Low-resource Languages 50%
• Four States 50%
• Transformer Model 50%
• Online Social Media 50%
• XLM-Roberta 50%
• Cyberbullying 50%
• Emoji Embedding 50%
• MBERT 50%
• Abusive Content 50%
• Bidirectional Gated Recurrent Unit (BiGRU) 50%
• Active Participation 50%
• IndicBERT 50%
• Handling Mechanism 50%
• Code-Mixed Language 50%
• Individual Languages 50%
• Model Fitting 50%
• Transformer Architecture 50%
• Indic 50%
• Computer-based Approach 50%
• Textual Content 50%
• Validation Loss 50%
• Individual Reputation 50%
• AUC Score 50%
• Training Loss 50%

Computer Science

• web user 100%
• Machine Learning 100%
• Learning System 100%
• Social Medium Platform 50%
• Language Modeling 50%
• Bidirectional Encoder Representations From Transformers 50%
• Textual Content 50%
• Individual Model 50%
• Handling Mechanism 50%
• Cyberbullying 50%
• Low-resource Languages 50%