Filtering objectionable information access based on click-through behaviours with deep learning methods

Keyphrases

• Access to Information 100%
• Bidirectional Long Short-term Memory (BiLSTM) 33%
• BiLSTM-CRF 66%
• Blocking Rate 33%
• Category Prediction 33%
• Click-through 100%
• Conditional Random Fields 33%
• Context-aware 33%
• Deep Learning Methods 100%
• Embedding Method 33%
• F1 Score 33%
• False Positive Rate 33%
• Large-scale Experiments 33%
• Online Information 33%
• Real Users 33%
• Real-time Filtering 33%
• Relative Improvement 66%
• User Access 33%
• User Categories 33%
• Web Surfing 33%

Engineering

• Deep Learning Method 100%
• Filtration 100%
• Long Short-Term Memory 100%
• Random Field 100%
• Real-Time Filtering 100%
• Scale Experiment 100%

Computer Science

• Bidirectional Long Short-Term Memory Network 100%
• Conditional Random Field 33%
• Deep Learning Method 100%
• False Positive Rate 33%
• Information Access 100%
• Long Short-Term Memory Neural Network 33%
• Online Information 33%
• Real-Time Filtering 33%
• Web Surfing 33%