Advances in Knowledge Discovery and Data Mining, Part II
16th Pacific-Asia Conference, PAKDD 2012, Kuala Lumpur, Malaysia, May 29-June 1, 2012, Proceedings, Part II
Published on 10. May 2012
XXIV, 445 pages
978-3-642-30220-6 (ISBN)
Description
The two-volume set LNAI 7301 and 7302 constitutes the refereed proceedings of the 16th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2012, held in Kuala Lumpur, Malaysia, in May 2012. The total of 20 revised full papers and 66 revised short papers were carefully reviewed and selected from 241 submissions. The papers present new ideas, original research results, and practical development experiences from all KDD-related areas. The papers are organized in topical sections on supervised learning: active, ensemble, rare-class and online; unsupervised learning: clustering, probabilistic modeling in the first volume and on pattern mining: networks, graphs, time-series and outlier detection, and data manipulation: pre-processing and dimension reduction in the second volume.
More details
Other editions
Additional editions
Pang-Ning Tan | Sanjay Chawla | Chin Kuan Ho
Advances in Knowledge Discovery and Data Mining, Part II
16th Pacific-Asia Conference, PAKDD 2012, Kuala Lumpur, Malaysia, May 29-June 1, 2012, Proceedings, Part II
Book
04/2012
Springer
€53.49
Shipment within 7-9 days
Persons
Content
- Title
- Preface
- Organization
- Table of Contents
- Pattern Mining: Networks, Graphs, Time-Series and Outlier Detection
- Heterogeneous Ensemble for Feature Drifts in Data Streams
- Introduction
- Related Work
- Proposed Framework
- Feature Selection Block
- Ensemble Block
- Experiments and Analysis
- Experimental Setup
- Experimental Results
- Conclusions
- References
- OMC-IDS: At the Cross-Roads of OLAP Mining and Intrusion Detection
- Introduction
- Scrutiny of the Related Work
- OMC-IDS: Intrusion Detection Based on Olap Mining and Classification
- Audit Data Cube: Construction and Manipulation
- Multidimensional Association Rule Mining
- Classification
- Experimental Results
- Conclusion and Perspectives
- References
- Towards Linear Time Overlapping Community Detection in Social Networks
- Introduction
- Related Work
- SLPA: Speaker-Listener Label Propagation Algorithm
- Tests in Synthetic Networks
- Methodology
- Identifying Overlapping Communities in LFR
- Identifying Overlapping Nodes in LFR
- Tests in Real-World Social Networks
- Identifying Overlapping Communities in Social Networks
- Identifying Overlapping Communities in Bipartite Networks
- Identifying Overlapping Nested Communities
- Conclusions
- References
- WeightTransmitter:Weighted Association Rule Mining Using LandmarkWeights
- Introduction
- Related Work
- Problem Definition
- Weight Transmitter Model
- Experimental Results
- Datasets
- Weight Estimation Evaluation
- Rule Evaluation
- Runtime Evaluation
- Conclusions
- References
- Co-occurring Cluster Mining for Damage Patterns Analysis of a Fuel Cell
- Introduction
- The Proposed Method: Co-occurring Cluster Mining
- Problems of the Conventional Methods
- The Requirements of a Co-occurrence Pattern
- The Objective Function
- The Algorithm
- Application to AE Data
- Damage Evaluation Test of Fuel Cells
- Division into Basket
- Calculation of Distance between AE Events
- The Design of the Object Function
- The Results of Extracted Damage Patterns
- Conclusion
- References
- New Exact Concise Representation of Rare Correlated Patterns: Application to Intrusion Detection
- Introduction and Motivations
- Basic Notions
- Characterization of the Rare Correlated Patterns
- Definition and Properties
- Characterization of the Rare Correlated Equivalence Classes
- The RcprMiner Algorithm
- Experimental Results
- Application to Intrusion Detection
- Description of the KDD 99 Dataset
- Summary of Experimentations and Discussion of Obtained Results
- Conclusion and Future Works
- References
- Life Activity Modeling of News Event on Twitter Using Energy Function
- Introduction
- Related Work
- Modeling Life Activity Using Energy Function
- Definition of Energy Function
- Energy of A Single Tweet
- Constant Growth and Decay
- Single-Pass Clustering with Energy Function
- Experiments and Evaluation
- Data Preparation
- Training Energy Transferred Factor and Decayed Factor
- News Event Detection Comparisons
- Conclusions
- References
- Quantifying Reciprocity in LargeWeighted Communication Networks
- Introduction
- Related Work
- Data Description
- Proposed Model: 3PL
- Comparison of 3PL to Competing Models
- Goodness of Fit
- 3PL at Work
- Reciprocity and Local Network Topology
- Weighted Reciprocity Metrics
- Reciprocity and Network Overlap
- Reciprocity and Degree Similarity
- Conclusions
- References
- Hierarchical Graph Summarization: Leveraging Hybrid Information through Visible and Invisible Linkage
- Introduction
- Related Work
- Basic Graph Summarization
- Hierarchical Graph Summarization
- Overview
- Incorporating Hierarchical Linkage
- Estimation of Document/Cluster Importance
- Experiments and Evaluation
- Dataset
- Evaluation Metrics
- Algorithms for Comparison
- Overall Performance Comparison
- Parameter Tuning
- Conclusions
- References
- Mining Mobile Users' Activities Based on Search Query Text and Context
- Introduction
- Related Work
- Methodology
- Data and Preprocessing
- Text and Context-Based User Activity Model
- Inference of Model
- Constrained TCUAM Model
- Experiment
- Data Set
- Experimental Setup
- Evaluation
- Results
- Case Study
- Conclusion
- References
- Spread of Information in a Social Network Using Influential Nodes
- Introduction
- Motivation
- Literature Review
- Maximizing Influence Spread
- Problem Definition
- Our Approach
- Detecting the core
- Experimental Results
- Conclusion
- References
- Discovering Coverage Patterns for Banner Advertisement Placement
- Introduction
- Model of Coverage Patterns
- Coverage Patterns
- Mining Coverage Patterns
- Coverage Pattern Extraction Algorithm
- Experimental Results
- Coverage Pattern Generation
- Scalability Experiment
- Usefulness of Coverage Patterns
- Conclusions and Future Work
- References
- Discovering Unknown But Interesting Items on Personal Social Network
- Introduction
- Related Works
- Social Networking
- Recommendation Systems
- Unknown But Interesting Recommendation System
- System Architecture
- Unknown But Interesting Algorithm
- Experiments
- Methodology
- Performance Evaluation
- Conclusions
- References
- The Pattern Next Door: Towards Spatio-sequential Pattern Discovery
- Introduction
- Related Work
- Spatio-sequential Patterns: Concepts and Definitions
- Preliminaries
- Spatio-sequential Patterns
- Spatio-temporal Participation
- Extraction of Spatio-sequential Patterns
- Experiments
- Conclusion and Perspectives
- References
- Accelerating Outlier Detection with Uncertain Data Using Graphics Processors
- Introduction
- Related Work
- Algorithm for Outlier Detection with Uncertain Data
- Serial and Parallel Implementations
- Serial Methods
- Parallel Methods
- Experimental Results
- Performance
- Quality
- Conclusion
- References
- Finding Collections of k-Clique Percolated Components in Attributed Graphs
- Introduction
- Pattern Definition
- Mining CoHoP Patterns
- Experiments
- Illustration of the Interest of the Patterns
- Performance Study
- Related Work
- Conclusion
- References
- Reciprocal and Heterogeneous Link Prediction in Social Networks
- Introduction
- Related Work
- Problem Statement
- Methods
- Feature Construction
- Learning and Testing
- Experiments
- Setup
- Results
- Conclusion
- References
- Detecting Multiple Stochastic Network Motifs in Network Data
- Introduction
- Network Motif Analysis in Social Media
- Canonical Forms of Subgraphs for Modeling Stochastic Motifs
- Finite Mixture Model
- Basic EM Algorithm
- Learning the Optimal Number of Motifs
- Experimental Results
- Results on Synthetic Networks
- Results on Benchmark Datasets
- Effectiveness of CEM2 in Estimating Optimal Number of Motifs
- Computational Complexity
- Conclusion and Future Works
- References
- Scalable Similarity Matching in Streaming Time Series
- Introduction and Motivations
- Key Notions
- Related Work
- The TriCons Algorithm
- Main Notions of the TriCons Algorithm
- Description of the TriCons Algorithm
- Experimental Results
- Conclusion and Future Work
- References
- Scalable Mining of Frequent Tri-concepts from Folksonomies
- Introduction and Motivations
- KeyNotions
- Related Work
- TheTRICONS Algorithm
- Main Notions of the TRICONS Algorithm
- Description of the TRICONS Algorithm
- Experimental Results
- Conclusion and Future Work
- References
- SHARD: A Framework for Sequential, Hierarchical Anomaly Ranking and Detection
- Introduction
- Related Literature
- Hierarchical Anomalies
- Anomaly Detection Framework
- Ontology Template
- Anomaly Tree Structure
- Baseline Anomaly Detectors
- Ranking Anomalies
- Anomaly Tree Visualization
- Empirical Evaluation
- Synthetic Data Experiments
- Event Attendance Data Results
- Climatology Data Results
- Stock Data Results
- Discussion
- Conclusions and Future Work
- References
- Instant Social Graph Search
- Introduction
- Problem Definition
- Algorithms
- Basic Ideas
- The Path Algorithm
- The Influence Algorithm
- The Diversity Algorithm
- Experimental Results
- Experiment Setup
- Accuracy Performance
- Analysis and Discussions
- Related Work
- Conclusions
- References
- Data Manipulation: Pre-processing and Dimension Reduction
- Peer Matrix Alignment: A New Algorithm
- Introduction
- Problem Definition
- The Proposed Algorithm
- Examples
- Experiments
- Related Works
- Conclusions
- References
- Domain Transfer Dimensionality Reduction via Discriminant Kernel Learning
- Introduction
- Brief Review of Prior Work
- Discriminant Multiple Kernel Learning
- Transfer Learning and Maximum Mean Discrepancy Formulation
- Semi-supervised Discriminant Analysis in Cross-Domain
- Standard Discriminant Kernel Learning Analysis
- Domain Transfer Kernel Learning for Discriminant Analysis
- Experiment
- Data Sets and Experiment Setup
- Experimental Results
- Conclusion
- References
- Prioritizing Disease Genes by Bi-Random Walk
- Introduction
- Methods
- Loss Function
- Bi-Random Walk
- Unbalanced Bi-Random Walk
- BiRW Algorithms
- Comparison of Random Walk Algorithms
- Experiments and Discussions
- Data Preparation
- Comparison with Other Methods
- Comparison of BiRW Variants
- Conclusion
- References
- Selecting Feature Subset via Constraint Association Rules
- Introduction
- Related Work
- Preliminaries
- Strong, Classification and Atomic Association Rules
- Definitions of Relevant, Redundant and Interactive Features
- Feature Subset Selection Algorithm
- FEAST Algorithm
- Experimental Results and Analysis
- Benchmark Data Sets
- Experimental Setup
- Results on the Synthetic Data Sets
- Results on the Real World Data Sets
- Conclusion
- References
- RadialViz: An Orientation-Free Frequent Pattern Visualizer
- Introduction
- Background and Related Work
- RadialViz: Our Proposed Visualizer
- Basic Representation of Frequent Patterns in RadialViz
- Other Features and Observations on RadialViz
- Evaluation
- Conclusions
- References
- Feature Weighting by RELIEF Based on Local Hyperplane Approximation
- Introduction
- The Proposed Method
- RELIEF
- Approximation by Local Hyperplane
- Experimental Results
- Selection of Classifier
- Fermat's Spiral Problem
- UCI Data Sets
- Discussion
- References
- Towards Identity Disclosure Control in Private Hypergraph Publishing
- Introduction
- Our Contributions
- Related Work
- Problem Statement
- Rank Attack
- Problem Definition
- Measuring Quality of Hypergraph Anonymization
- Algorithms
- Rank Anonymization
- Hypergraph Construction
- Experiments
- Rank Attack on Real-World Data
- Impact on Anonymizing Cost ZA
- Impact on Information Loss
- Conclusion
- References
- EWNI: Efficient Anonymization of Vulnerable Individuals in Social Networks
- Introduction
- Related Literature
- Graph Structure and Privacy Model
- Background
- PrivacyModel
- Anonymization Algorithms
- FindingWeak Nodes and Neighborhood Subgraphs
- Anonymizing G
- Experiments
- Conclusions
- References
- A Pruning-Based Approach for Searching Preciseand Generalized Region for Synthetic Minority Over-Sampling
- Introduction
- Related Work
- Methodology
- TRIM Criteria
- TRIM Algorithm
- Experimental Results
- Conclusions
- References
- Towards More Efficient Multi-label Classification Using Dependent and Independent Dual Space Reduction
- Introduction
- Preliminaries
- Definition of Multi-label Classification Task
- Singular Value Decomposition (SVD)
- Two Proposed Approaches
- Dependent Dual Space Reduction (DDSR)
- Independent Dual Space Reduction (IDSR)
- Threshold Selection
- Datasets and Experimental Settings
- Experimental Results
- Conclusion
- References
- Automatic Identification of Protagonist in Fairy Tales Using Verb
- Introduction and Motivation
- Technologies Background
- Stanford Parser
- WordNet
- System Framework
- Experiments and Discussions
- Dataset
- Results and Discussion
- Conclusion
- References
- CD: A Coupled Discretization Algorithm
- Introduction
- Related Work
- Problem Statement
- Discretization Algorithm Based on Information Entropy
- Discretization Algorithm Based on Positive Domain
- Discretization Algorithm Based on the Coupling
- Experiment and Evaluation
- Internal Criteria Comparison
- External Criterion Comparison
- Conclusion
- References
- Co-embedding of Structurally Missing Data by Locally Linear Alignment
- Introduction
- Related Works
- Problem Definition
- Locally Linear Alignment Co-embedding
- Basic Idea
- Unsupervised Locally Linear Alignment Co-embedding
- Regularization
- Semi-supervised Co-embedding
- Experiment
- Experiment 1: Structure from Motion Task
- Experiment 2: Mapping and Localization forWireless Devices
- Conclusion
- References
- Relevant Feature Selection from EEG Signal for Mental Task Classification
- Introduction
- Feature Extraction from EEG
- Empirical Mode Decomposition (EMD)
- Estimation of Various Parameters
- Feature Selection
- Experimental Set-Up and Results
- Conclusion
- References
- Author Index
