Method Term: count-based feature reduction

To view the usage of count-based feature reduction in the ACL ARC 2.0 corpus click here (i.e., KWIC using NoSke).

Terms semantically related to count-based feature reduction:

• layer training
• single layer training
• boundary event detection
• tagging-like metadata detection
• transcript evaluation
• telephone shopping
• unlabeled instance selection
• keyword-driven speech recognition
• manual thematic role assignment
• verb class discovery
• monolingual text generation
• fixed parsing
• example-based classification

Note that only terms are manually validated, not their relationships. See the ACL RD-TEC 1.0 documentations.