RoBERTa-LSTM: A Hybrid Model for Sentiment Analysis With Transformer and Recurrent Neural Network

Computer Science

• Sentiment Analysis 100%
• Recurrent Neural Network 100%
• Long Short-Term Memory Network 100%
• Sequence Model 66%
• Word Embedding 33%
• Data Augmentation 33%
• Bidirectional Encoder Representations From Transformers 33%
• Decision-Making 16%
• Social Media 16%
• Experimental Result 16%
• Deep Learning 16%
• Rapid Development 16%
• Computation Time 16%
• Execution Time 16%
• Memory Model 16%
• Minority Class 16%
• Social Medium Technology 16%
• Embedding Space 16%

Keyphrases

• Transformer Neural Network 100%
• Text Sentiment Analysis 20%
• Word Embedding Space 20%
• Word Vector 20%
• IMDb Dataset 20%
• GloVe Embedding 20%
• Lexical Diversity 20%
• Similar Words 20%