Principles of Data Mining and Knowledge Discovery

This volume contains papers selected for presentation at PKDD’2000, the Fourth European Conference on Principles and Practice of Knowledge Discovery in - tabases. The rst meeting was held in Trondheim, Norway, in June 1997, the second in Nantes, France, in September 1998, and the third in Prague, Czech Republic, in September 1999. PKDD 2000 was organized in Lyon, France, on 13{16 September 2000. The conference was hosted by the Equipe de Recherche en Ing enierie des Conna- sances at the Universit e Lumi ere Lyon 2. We wish to express our thanks to the sponsors of the Conference, to the University Claude Bernard Lyon 1, the INSA of Lyon, the Conseil g en eral of the R^one, the R egion Rh^one Alpes, SPSS France, AFIA, and the University of Lyon 2, for their generous support. Knowledge discovery in databases (KDD), also known as data mining, p- vides tools for turning large databases into knowledge that can be used in pr- tice. KDD has been able to grow very rapidly since its emergence a decade ago by drawing its techniques and data mining experiences from a combination of many existing research areas: databases, statistics, mathematical logic, machine learning, automated scienti c discovery, inductive logic programming, arti cial intelligence, visualization, decision science, knowledge management, and high performance computing. The strength of KDD came initially from the value - ded by the creative combination of techniques from the contributing areas.

Towards Broader Foundations.- Multi-relational Data Mining, Using UML for ILP.- An Apriori-Based Algorithm for Mining Frequent Substructures from Graph Data.- Basis of a Fuzzy Knowledge Discovery System.- Rules and Trees.- Confirmation Rule Sets.- Contribution of Dataset Reduction Techniques to Tree-Simplification and Knowledge Discovery.- Combining Multiple Models with Meta Decision Trees.- Databases and Reward-Based Learning.- Materialized Data Mining Views.- Approximation of Frequency Queries by Means of Free-Sets.- Application of Reinforcement Learning to Electrical Power System Closed-Loop Emergency Control.- Efficient Score-Based Learning of Equivalence Classes of Bayesian Networks.- Classication.- Quantifying the Resilience of Inductive Classification Algorithms.- Bagging and Boosting with Dynamic Integration of Classifiers.- Zoomed Ranking: Selection of Classification Algorithms Based on Relevant Performance Information.- Some Enhancements of Decision Tree Bagging.- Association Rules and Exceptions.- Relative Unsupervised Discretization for Association Rule Mining.- Mining Association Rules: Deriving a Superior Algorithm by Analyzing Today’s Approaches.- Unified Algorithm for Undirected Discovery of Exception Rules.- Sampling Strategies for Targeting Rare Groups from a Bank Customer Database.- Instance-Based Discovery.- Instance-Based Classification by Emerging Patterns.- Context-Based Similarity Measures for Categorical Databases.- A Mixed Similarity Measure in Near-Linear Computational Complexity for Distance-Based Methods.- Fast Feature Selection using Partial Correlation for Multi-valued Attributes.- Clustering and Classification.- Fast Hierarchical Clustering Based on Compressed Data and OPTICS.- Accurate Recasting of Parameter Estimation Algorithms Using Sufficient Statistics for Efficient Parallel Speed-Up.- Predictive Performance of Weighted Relative Accuracy.- Quality Scheme Assessment in the Clustering Process.- Time Series.- Algorithm for Matching Sets of Time Series.- MSTS: A System for Mining Sets of Time Series.- Learning First Order Logic Time Series Classifiers: Rules and Boosting.- Posters.- Learning Right Sized Belief Networks by Means of a Hybrid Methodology.- Algorithms for Mining Share Frequent Itemsets Containing Infrequent Subsets.- Discovering Task Neighbourhoods through Landmark Learning Performances.- Induction of Multivariate Decision Trees by Using Dipolar Criteria.- Inductive Logic Programming in Clementine.- A Genetic Algorithm-Based Solution for the Problem of Small Disjuncts.- Clustering Large, Multi-level Data Sets: An Approach Based on Kohonen Self Organizing Maps.- Trees and Induction Graphs for Multivariate Response.- CEM - A Program for Visualization and Discovery in Email.- Image Access and Data Mining: An Approach.- Decision Tree Toolkit: A Component-Based Library of Decision Tree Algorithms.- Determination of Screening Descriptors for Chemical Reaction Databases.- Prior Knowledge in Economic Applications of Data Mining.- Temporal Machine Learning for Switching Control.- Improving Dissimilarity Functions with Domain Knowledge, applications with IKBS system.- Mining Weighted Association Rules for Fuzzy Quantitative Items.- Centroid-Based Document Classification: Analysis and Experimental Results.- Applying Objective Interestingness Measures in Data Mining Systems.- Observational Logic Integrates Data Mining Based on Statistics and Neural Networks.- Supporting Discovery in Medicine by Association Rule Mining of Bibliographic Databases.- Collective Principal Component Analysis from Distributed, Heterogeneous Data.- Hierarchical Document Clustering Based on Tolerance Rough Set Model.- Application of Data-Mining and Knowledge Discovery in Automotive Data Engineering.- Towards Knowledge Discovery from cDNA Microarray Gene Expression Data.- Mining with Cover and Extension Operators.- A User-Driven Process for Mining Association Rules.- Learning from Labeled and Unlabeled Documents: A Comparative Study on Semi-Supervised Text Classification.- Schema Mining: Finding Structural Regularity among Semistructured Data.- Improving an Association Rule Based Classifier.- Discovery of Generalized Association Rules with Multiple Minimum Supports.- Learning Dynamic Bayesian Belief Networks Using Conditional Phase-Type Distributions.- Discovering Differences in Patients with Uveitis through Typical Testors by Class.- Web Usage Mining: How to Efficiently Manage New Transactions and New Clients.- Mining Relational Databases.- Interestingness in Attribute-Oriented Induction (AOI): Multiple-Level Rule Generation.- Discovery of Characteristic Subgraph Patterns Using Relative Indexing and the Cascade Model.- Transparency and Predictive Power.- Clustering Distributed Homogeneous Datasets.- Empirical Evaluation of Feature Subset Selection Based on a Real-World Data Set.- Discovery of Ambiguous Patterns in Sequences Application to Bioinformatics.- Action-Rules: How to Increase Profit of a Company.- Aggregation and Association in Cross Tables.- An Experimental Study of Partition Quality Indices in Clustering.- Expert Constrained Clustering: A Symbolic Approach.- An Application of Association Rules Discovery to Geographic Information Systems.- Generalized Entropy and Projection Clustering of Categorical Data.- Supporting Case Acquisition and Labelling in the Context of Web Mining.- Indirect Association: Mining Higher Order Dependencies in Data.- Discovering Association Rules in Large, Dense Databases.- Providing Advice to Website Designers Towards Effective Websites Re-organization.- Clinical Knowledge Discovery in Hospital Information Systems: Two Case Studies.- Knowledge Discovery Using Least Squares Support Vector Machine Classifiers: A Direct Marketing Case.- Lightweight Document Clustering.- Automatic Category Structure Generation and Categorization of Chinese Text Documents.- Mining Generalized Multiple-level Association Rules.- An Efficient Approach to Discovering Sequential Patterns in Large Databases.- Using Background Knowledge as a Bias to Control the Rule Discovery Process.