Learning to Classify Documents with Only a Small Positive Training Set.
Xiaoli LiBing LiuSee-Kiong NgPublished in: ECML (2007)
Keyphrases
- learning to classify
- training set
- text classification
- information retrieval
- training data
- document collections
- text documents
- relevant documents
- test set
- classification accuracy
- document classification
- document retrieval
- xml documents
- small number
- data sets
- metadata
- training examples
- information retrieval systems
- positive and negative
- training samples
- document analysis
- document representation
- web documents
- feature space
- cross validation
- classification algorithm
- document clustering
- test data
- database
- learning algorithm
- feature selection
- digital libraries
- active learning
- text categorization
- nearest neighbor