Tracking Concept Drift at Feature Selection Stage in SpamHunting: An Anti-spam Instance-Based Reasoning System.
José Ramon MéndezFlorentino Fernández RiverolaEva Lorenzo IglesiasFernando DíazJuan M. CorchadoPublished in: ECCBR (2006)
Keyphrases
- concept drift
- anti spam
- anti spam filtering
- feature selection
- data streams
- spam filtering
- ensemble classifier
- drifting concepts
- non stationary
- drift detection
- data stream classification
- change detection
- spam detection
- classification algorithm
- data distribution
- spam emails
- text categorization
- multi class
- text classification
- feature extraction
- remote sensing
- support vector
- machine learning
- web spam
- spam filters
- neural network
- data structure
- classification accuracy
- feature subset
- naive bayes
- feature set
- image processing
- dimensionality reduction
- support vector machine