Principles of Data Mining and Knowledge Discovery

This book constitutes the refereed proceedings of the 6th European Conference on Principles of Data Mining and Knowledge Discovery, PKDD 2002, held in Helsinki, Finland in August 2002.The 39 revised full papers presented together with 4 invited contributions were carefully reviewed and selected from numerous submissions. Among the topics covered are kernel methods, probabilistic methods, association rule mining, rough sets, sampling algorithms, pattern discovery, web text mining, meta data clustering, rule induction, information extraction, dependency detection, rare class prediction, classifier systems, text classification, temporal sequence analysis, unsupervised learning, time series analysis, medical data mining, etc.

Inhalt

Contributed Papers.- Optimized Substructure Discovery for Semi-structured Data.- Fast Outlier Detection in High Dimensional Spaces.- Data Mining in Schizophrenia Research - Preliminary Analysis.- Fast Algorithms for Mining Emerging Patterns.- On the Discovery of Weak Periodicities in Large Time Series.- The Need for Low Bias Algorithms in Classification Learning from Large Data Sets.- Mining All Non-derivable Frequent Itemsets.- Iterative Data Squashing for Boosting Based on a Distribution-Sensitive Distance.- Finding Association Rules with Some Very Frequent Attributes.- Unsupervised Learning: Self-aggregation in Scaled Principal Component Space.- A Classification Approach for Prediction of Target Events in Temporal Sequences.- Privacy-Oriented Data Mining by Proof Checking.- Choose Your Words Carefully: An Empirical Study of Feature Selection Metrics for Text Classification.- Generating Actionable Knowledge by Expert-Guided Subgroup Discovery.- Clustering Transactional Data.- Multiscale Comparison of Temporal Patterns in Time-Series Medical Databases.- Association Rules for Expressing Gradual Dependencies.- Support Approximations Using Bonferroni-Type Inequalities.- Using Condensed Representations for Interactive Association Rule Mining.- Predicting Rare Classes: Comparing Two-Phase Rule Induction to Cost-Sensitive Boosting.- Dependency Detection in MobiMine and Random Matrices.- Long-Term Learning for Web Search Engines.- Spatial Subgroup Mining Integrated in an Object-Relational Spatial Database.- Involving Aggregate Functions in Multi-relational Search.- Information Extraction in Structured Documents Using Tree Automata Induction.- Algebraic Techniques for Analysis of Large Discrete-Valued Datasets.- Geography of Di.erences between Two Classes of Data.- Rule Induction for Classification of Gene Expression Array Data.- Clustering Ontology-Based Metadata in the Semantic Web.- Iteratively Selecting Feature Subsets for Mining from High-Dimensional Databases.- SVM Classification Using Sequences of Phonemes and Syllables.- A Novel Web Text Mining Method Using the Discrete Cosine Transform.- A Scalable Constant-Memory Sampling Algorithm for Pattern Discovery in Large Databases.- Answering the Most Correlated N Association Rules Efficiently.- Mining Hierarchical Decision Rules from Clinical Databases Using Rough Sets and Medical Diagnostic Model.- Efficiently Mining Approximate Models of Associations in Evolving Databases.- Explaining Predictions from a Neural Network Ensemble One at a Time.- Structuring Domain-Specific Text Archives by Deriving a Probabilistic XML DTD.- Separability Index in Supervised Learning.- Invited Papers.- Finding Hidden Factors Using Independent Component Analysis.- Reasoning with Classifiers.- A Kernel Approach for Learning from Almost Orthogonal Patterns.- Learning with Mixture Models: Concepts and Applications.