Co-training using prosodic and lexical information for sentence segmentation.
Ümit GüzSébastien CuendetDilek Hakkani-TürGökhan TürPublished in: INTERSPEECH (2007)
Keyphrases
- co training
- lexical information
- semantic roles
- part of speech
- semi supervised
- single view
- multi view
- semi supervised learning
- unlabeled data
- text classification
- supervised learning
- parse tree
- language processing
- natural language
- semantic role labeling
- semantic analysis
- labeled data
- wordnet
- information retrieval
- natural language processing
- named entities
- word sense disambiguation
- co occurrence
- active learning
- object recognition
- training data