Enhancing sarcasm detection in sentiment analysis for cyberspace safety using advanced deep learning techniques.

Area of Science:

• Computer Science
• Artificial Intelligence
• Natural Language Processing

Background:

• Social media platforms are widely used for communication but also facilitate cyberbullying.
• Detecting sarcasm in online content is crucial for identifying and mitigating cyberbullying.
• Existing automated systems struggle with the nuances of sarcasm detection.

Purpose of the Study:

• To develop an advanced deep learning model for accurate sarcasm detection in social media.
• To enhance the mitigation of cyberbullying by improving automated detection capabilities.
• To validate the model's effectiveness using benchmark datasets and established metrics.

Main Methods:

• A deep learning approach combining a Convolutional Neural Network (CNN) for feature extraction and an Attention Mechanism-based Bidirectional Long Short-Term Memory with Gated Recurrent Unit (AM-BLSTM-GRU) for prediction.
• Utilized sarcasm detection datasets from Kaggle and News Headlines, incorporating standard NLP-based auxiliary features.
• Employed the Enhanced Sinogramic Red Deer (ESRD) optimizer for effective classifier parameter optimization.

Main Results:

• The proposed AM-BLSTM-GRU model demonstrated superior performance in sarcasm detection and sentiment classification compared to existing deep learning methods.
• The model achieved high accuracy on popular benchmark datasets and evaluation metrics.
• The approach proved effective in identifying and reducing harmful online behaviors associated with cyberbullying.

Conclusions:

• The developed deep learning model offers a robust solution for detecting cyberbullying through sarcasm analysis.
• This method significantly improves the accuracy and efficiency of identifying harmful online content.
• The findings contribute to creating safer online environments by mitigating cyberbullying effectively.