An LSTM approach to patent classification based on fixed hierarchy vectors

Keyphrases

• Classification Basis 100%
• Classification Methods 100%
• Classification Performance 50%
• Document Representation 50%
• Document Structure 100%
• Embedding Algorithm 50%
• Fixed Structure 50%
• Innovative Technique 50%
• International Patent Classification 50%
• Latent Dirichlet Allocation Algorithm 50%
• Local Context 50%
• LSTM Approach 100%
• Multiple Levels 50%
• New Classification 50%
• Paragraph Vector 50%
• Patent Classification 100%
• Patent Data 50%
• Retrieval Methods 50%
• Sequential Classification 50%
• Text Classification 50%
• Text Collection 50%
• Text Processing 50%
• Text Retrieval 50%

Computer Science

• Application Area 50%
• Classification Method 100%
• Classification Performance 50%
• Document Representation 50%
• Embedding Algorithm 50%
• Latent Dirichlet Allocation 50%
• Long Short-Term Memory Network 100%
• retrieval method 50%
• Taxonomy Classification 50%
• Text Classification 50%
• Text Collection 50%
• Text Processing 50%