Automatic Text Classification课件.ppt

Multi-Topic Text Classification ?2007 Yutaka Sasaki, University of Manchester 精品文档 Examples of information filtering web documents ?2007 Yutaka Sasaki, University of Manchester 精品文档 Examples of information extraction web documents about accidents Date: 04/12/03 Place: London Type: traffic Casualty: 5 Key information on accidents ?2007 Yutaka Sasaki, University of Manchester 精品文档 Examples of text classification web documents ?2007 Yutaka Sasaki, University of Manchester sports economics 精品文档 Text Classification Applications E-mail spam filtering Categorize newspaper articles and newswires into topics Organize Web pages into hierarchical categories Sort journals and abstracts by subject categories (e.g., MEDLINE, etc.) Assigning international clinical codes to patient clinical records ?2007 Yutaka Sasaki, University of Manchester 精品文档 Simple text classification example You want to classify documents into 4 classes: economics, sports, science, life. There are two approaches that you can take: rule-based approach write a set of rules that classify documents machine learning-based approach using a set of sample documents that are classified into the classes (training data), automatically create classifiers based on the training data ?2007 Yutaka Sasaki, University of Manchester 精品文档 Comparison of Two Approaches (1) Rule-based classification Pros: very accurate when rules are written by experts classification criteria can be easily controlled when the number of rules are small. Cons: sometimes, rules conflicts each other maintenance of rules becomes more difficult as the number of rules increases the rules have to be reconstructed when a target domain changes low coverage because of a wide variety of expressions ?2007 Yutaka Sasaki, University of Manchester 精品文档 Comparison of Two Approaches (2) Machine Learning-based approach Pros: domain independent high predictive performance Cons: not accountable for classification results training data required