Advanced Data Mining and Applications

Name: Advanced Data Mining and Applications | 20th International Conference, ADMA 2024, Sydney, NSW, Australia, December 3-5, 2024, Proceedings, Part III
Brand: Springer
Price: 85.59 EUR
Availability: OnlineOnly

20th International Conference, ADMA 2024, Sydney, NSW, Australia, December 3-5, 2024, Proceedings, Part III

Quan Z. Sheng Gill Dobbie Jing Jiang Xuyun Zhang Wei Emma Zhang Yannis Manolopoulos Jia Wu Wathiq Mansoor Congbo Ma(Editor)

Springer (Publisher)

Published on 14. December 2024

XIX, 448 pages

E-Book

PDF with digital watermarking

System requirements

978-981-96-0821-8 (ISBN)

€85.59incl. 7% vat

System requirements

for PDF with digital watermarking

E-Book Single Licence

Available for download

Description

More details

Other editions

Content

Intro
Preface
Organisation
Contents - Part III
Graph Mining
Verifiable Graph-Based Approximate Nearest Neighbor Search
1 Introduction
2 Related Works
2.1 Graph-Based Approximate Nearest Neighbor Search
2.2 Verifiable Nearest Neighbor Search
3 Preliminaries
3.1 Hierarchical Clustering-Based Nearest Neighbor Graph
3.2 Guided Tree
3.3 Merkle Hash Tree (MHT)
3.4 The Threat Model
4 Our Scheme
4.1 Initialization Phase
4.2 Query Processing Phase
4.3 Verification Phase
5 Security Discussion
6 Experiments
6.1 Setup
6.2 Impact of k and Number of Queries on VO Size
6.3 Computational Overhead
7 Conclusion
References
Depth-Enhanced Contrast Attribute Graph Clustering
1 Introduction
2 Related Work
2.1 Deep Graph Clustering
2.2 Graph Data Augmentation
3 Method
3.1 Notations
3.2 Deep Enhancement Module
3.3 Contrast Learning Module
3.4 Self-optimizing Module
3.5 Overall Objective
4 Experiments
4.1 Baseline Dataset
4.2 Baseline
4.3 Experimental Setup
4.4 Evaluation Metrics
4.5 Performance Comparison
4.6 Ablation Experiment
4.7 Sensitivity Analysis
4.8 Visualization
4.9 Conclusion
References
FCMH: Fast Cluster Multi-hop Model for Graph Fraud Detection
1 Introduction
2 Related Work
2.1 Graph Neural Network
2.2 Graph Fraud Detection
3 Methodology
3.1 Problem Formulation
3.2 Random Cluster Subgraph Division
3.3 Multi-hop Neighbor Difference Aggregation
3.4 Downsampling and Optimization Objective
4 Experiments
4.1 Experiment Settings
4.2 Classification Performance
4.3 Time Efficiency
4.4 Ablation Study
5 Conclusion
References
Emotion Graph Augmentation for Detecting Fake News in Online Social Networks
1 Introduction
2 Related Work
2.1 Text Based Methods
2.2 Graph Based Methods
2.3 Emotion Based Methods
3 Problem Statement
4 Methodology
4.1 Semantic Graph Construction
4.2 Emotion Graph Construction
4.3 Graph Augmentation with Adversarial Perturbations
4.4 Propagation of Semantics and Emotions
4.5 Fake News Detection
5 Experiments
5.1 Experimental Settings
5.2 Overall Performance
5.3 Ablation Study
5.4 Parameter Analysis
5.5 Case Study
6 Conclusion
References
BiF-AC: A Bidirectional Feedback Actor-Critic Framework for UAV-UGV Graph-Based Search and Rescue Operations
1 Introduction
2 UAV-UGV Coordination System Model
2.1 System Model
2.2 Problem Formulation
3 The Proposed Method
3.1 The Principles of Actor-Critic
3.2 The Bidirectional Feedback Actor-Critic Algorithm
4 Empirical Studies
4.1 Experimental Settings
4.2 Performance Evaluation
5 Conclusion
References
RWEM: An In-Memory Random Walk Based Node Embedding Framework on Multiplex User-Item Graphs
1 Introduction
2 Background and Related Work
2.1 Graph Theory
2.2 Stochastic Markov Process
2.3 Node Embedding
3 RWEM Framework
3.1 Embedding Input
3.2 Autocovariance-Based Similarity
4 Evaluation
4.1 Environment
4.2 Setup
4.3 Results
5 Conclusion
References
Feature-Aware Unsupervised Detection of Important Nodes in Graphs
1 Introduction
2 Related Works
3 Preliminaries
3.1 Problem Statement
3.2 Graph Convolutional Networks
4 Proposed Model
4.1 Feature-Aware Personalized PageRank
4.2 Model Architecture
4.3 Training Time Cost Analysis
5 Experiments
5.1 Node Classification
5.2 Active Learning
6 Conclusion and Future Work
References
HHP: A Hybrid Partitioner for Large-Scale Hypergraph
1 Introduction
2 Related Work
3 Preliminaries
4 Hybrid Hypergraph Partitioner
4.1 Basic Algorithm
4.2 Improved Online Partition Algorithm
4.3 Hybrid Hypergraph Partitioner
5 Evaluation
5.1 Experimental Setup
5.2 Hypergraph Partitioning
5.3 Experimental on MinMax++
5.4 Study on Hybrid Strategies
6 Conclusions
References
Graph Fusion Based Autoencoder for Node Clustering
1 Introduction
2 Related Work
2.1 Deep Clustering
2.2 Autoencoder
3 Method
3.1 Graph Fusion
3.2 Representation Learning
3.3 Node Clustering
4 Experiments
4.1 Experimental Setting
4.2 Result Analysis
4.3 Ablation Study
4.4 Parameter Sensitivity Analysis
5 Conclusion
References
Regional Food Culture Preference Mining Based on Restaurant POI
1 Introduction
2 Related Work
3 Dataset Construction
4 Methods
4.1 Statistical Analysis
4.2 Community Detection
5 Study of Chinese Cuisines
5.1 Data Distribution
5.2 Geographical Factors
5.3 Economical Factors
5.4 Population Factors
6 Clustering Analysis
6.1 Experimental Setup
6.2 Overall Performance
6.3 Ablation Study
6.4 Hyperparameter Analysis
6.5 Visualization Analysis
7 Conclusions
References
Multi-task Learning of Heterogeneous Hypergraph Representations in LBSNs
1 Introduction
2 Model and Problem Formulation
3 Constructing the Heterogeneous Hypergraph
4 Heterogeneous Hypergraph Learning
4.1 Hypergraph Input
4.2 Adaptive Heterogeneous Hypergraph Convolutional Network
5 Multi-task Learning
6 Empirical Evaluation
6.1 End-to-End Comparison
6.2 Ablation Testing
6.3 Hyperparameter Sensitivity
7 Conclusion
References
Graph Contrastive Learning for Dissolved Gas Analysis
1 Introduction
2 Preliminaries
2.1 Notation
2.2 Constructing KNN Graph
3 Methodology
3.1 Dual-Channel Graph Representation Learning
3.2 Ranking Contrastive Learning
3.3 Fault Detection
4 Experiment
4.1 Experimental Setup(RQ1)
4.2 Performance Comparison
4.3 Ablation Study(RQ2)
4.4 Parameter Analysis(RQ3)
5 Conclusion
References
GCS: A Graph-Augmented Semi-supervised Contrastive Learning Approach for Imbalanced Dissolved Gas Analysis in Power Transformers
1 Introduction
2 Preliminaries
2.1 Notations
2.2 Imbalance Settings and Problem Definition
3 Methodology
3.1 Graph Construction
3.2 Semi-supervised Contrastive Learning
3.3 Graph Augmentation
3.4 Classification and Model Optimization
4 Experiments
4.1 Experiment Settings
4.2 Main Comparison Results (RQ1)
4.3 Ablation Study (RQ2)
4.4 Imbalance Comparison(RQ3)
5 Conclusion
References
Contrastive Learning Based on Bipartite Graphs for Interpretable Knowledge Tracing
1 Introduction
2 Related Work
3 Methodology
3.1 Problem Definition
3.2 Model Overview
3.3 Embedding and Knowledge Structure
3.4 Bipartite Graph Attention Network
3.5 Bipartite Graph Contrastive Learning
3.6 Prediction
3.7 Model Optimization
4 Experimental Results
4.1 Datasets
4.2 Baselines and Experimental Settings
4.3 Implementation Details
4.4 Performance Analysis
4.5 Interpretability Discussion
4.6 Ablation Study
4.7 Conclusion
References
Graph Data Understanding and Interpretation Enabled by Large Language Models
1 Introduction
2 Method
2.1 Preliminary
2.2 Heterogeneous Data Representation Learning
2.3 Converter Alignment Tuning
2.4 Retrieval Augmented Thoughts
3 Experiments
3.1 Training Details
3.2 Baseline Method
3.3 Performance Comparison
3.4 Ablation Experiments
4 Conclusion
References
SDM-GAT: StylisticFP Detection Method Based on Graph Attention Network
1 Introduction
2 Background and Related Work
2.1 Tracking Development
2.2 Detection Methods
3 Methodology
3.1 Preliminaries
3.2 Graph Building
3.3 Graph Attention Network
4 Experiment
4.1 Datasets
4.2 Baselines and Metric
4.3 Implementation Details
4.4 Baseline Model Comparison
4.5 Impact of Graph Pruning
5 Conclusion
References
Anomaly Aligned Subgraphs Detection on Multi-layer Attributed Networks
1 Introduction
2 Related Work
2.1 Anomaly Detection
2.2 Network Alignment
3 Methodology
3.1 Anomaly Detection
3.2 Network Alignment
3.3 Update Anomaly Subgraph Node Set
4 Experiment
4.1 Experiment Settings
4.2 Results
4.3 Case Study
5 Conclusion
References
Path-Aware Siamese Graph Neural Network for Link Prediction
1 Introduction
2 Related Work
3 Method
3.1 Problem Formulation
3.2 Model Buildup
3.3 Contrastive Learning
4 Experiments
4.1 Datasets and Task
4.2 Baselines
4.3 Metrics and Settings
4.4 Abalation Study
5 Conclusion
References
GEM-GNN: Group Enhanced Multi-relation Graph Neural Networks for Fraud Detection
1 Introduction
2 Related Work
3 Proposed Model
3.1 Problem Definition
3.2 Model Architecture
3.3 Neighbor Aggregation Module
3.4 Group-Based Aggregation Module
3.5 Optimization
4 Experiments
4.1 Experimental Settings
4.2 Performance Comparison
4.3 Sensitivity Analysis
4.4 Training Process Study
5 Application
6 Conclusion
References
Spatial Data Mining
ESNet: Perceptive Spatial-Spectral Fusion with Multi-stage Reconstruction for Pansharpening
1 Introduction
2 The Proposed Model
2.1 ESNet
2.2 Enhanced Spatial Spectral Attention Module
2.3 Multi-scale Reconstruction Module
2.4 Loss Function
3 Experiment
3.1 Datasets and Settings
3.2 Accuracy Evaluation
3.3 Results
3.4 Ablation Experiment
4 Discussion
5 Conclusion
References
Towards Unified Spatio-Temporal Index for Hybrid Trajectory Search
1 Introduction
2 Preliminaries
2.1 Trajectory Data
2.2 Trajectory Queries
3 Unified Spatio-Temporal Index
3.1 Real-Time Grids
3.2 Spatio-Temporal Cubes
3.3 Lightweight Posting Lists
4 Hybrid Trajectory Search Algorithm
4.1 Accelerating Range Queries
4.2 Accelerating kNN Queries
5 Experiments
5.1 Experimental Setup
5.2 Efficiency of Range Query
5.3 Efficiency of kNN Query
6 Related Work
7 Conclusion and Future Work
References
VQPulsar: Pulsar Candidate Analysis via Deep Generative Model
1 Introduction
2 Related Work
3 The Proposed VQPulsar Model
3.1 Overview
3.2 Construction of Diagnostic Feature Figures
3.3 Representation Learning with VQVAE
3.4 Latent Space Analysis
3.5 Data Augmentation with GPT-VQVAE
4 Experimental Settings
4.1 Dataset
4.2 Settings
4.3 Evaluation Metrics
5 Experimental Results
5.1 Latent Space Analysis
5.2 Augmentation Analysis
6 Conclusion
References
Efficient Shortest Time Query in Public Transportation Networks
1 Introduction
2 Related Work
3 Problem Definition
4 Our Solution
4.1 Timetable Graph Tree Decomposition
4.2 TT-H2H Index Construction
4.3 Timetable Shortest Time Query
5 Experimental Evaluation
5.1 Experiment Setting
5.2 Experiment Results
6 Conclusion
References
When Road Networks Make a Difference: User Identity Linkage with Trajectory Data
1 Introduction
2 Related Work
3 Preliminaries
4 Overview of UIL-RN
5 Proposed Model UIL-RN
5.1 Graph Attention Network Module with a Transfer Probability Matrix (GAT-TPM)
5.2 Transformer-Based Encoder
5.3 Matcher
6 Experiment Study
6.1 Dataset
6.2 Parameter Setting and Evaluation Metrics
6.3 Compared Methods
6.4 Experimental Results
6.5 Ablation Study
6.6 Analysis of Parameters
7 Conclusion
References
A Transformer Based Malicious Traffic Detection Method in Android Mobile Networks
1 Introduction
2 Related Works
2.1 Machine Learning Based Works
2.2 Deep Learning Based Works
3 Methodology
3.1 Data Processing Method
3.2 Transformer for Detection
4 Experiment and Evaluation
4.1 Two-Classification Scenario
4.2 Five-classification Scenario
5 Conclusion
References
P2S-Sketch: A Sketch Family for Priority-Aware Per-Flow Spread Measurement in Network Data Stream
1 Introduction
2 Related Work
2.1 Priority-Agnostic Sketches
2.2 Priority-Aware Sketches
3 Design of P2S-Sketch Family
3.1 Generic Data Structure of P2S-Sketch Family
3.2 Generic Operation of P2S-Sketch Family
3.3 P2S-Basic
3.4 P2S-DT
3.5 P2S-FC
3.6 Performance Analysis
4 Experiments
4.1 Experimental Setup
4.2 Parameter Evaluation
4.3 Accuracy
4.4 Speed
4.5 Key Observations
5 Conclusion
References
STA: Enhancing Spatio-temporal Crowd Flow Prediction Using Attention-based Deep Learning and Feature Similarity
1 Introduction
2 Related Work
3 Problem Formulation
3.1 Region
3.2 Inflow/Outflow
3.3 Crowd Flow Prediction
4 Our Model
4.1 Slice Into Patches
4.2 Embedding
4.3 Attention Computation
5 Experiment
5.1 Datasets
5.2 Evaluation Metric
5.3 Results on TaxiBJ
5.4 Results on BikeNYC
5.5 The Analysis of Experimental Results
5.6 The Impact of Data Volume on RMSE
5.7 Spatio-Temporal Features Test
6 Conclusion
References
Spatial-Temporal Mamba Network for EEG-Based Motor Imagery Classification
1 Introduction
2 Related Works
3 Methodology
3.1 Embedding Component
3.2 Mamba Encoders
3.3 Classifier
4 Experiment and Analysis
4.1 Datasets
4.2 Experimental Setup and Baseline Models
4.3 Comparison of Performance
4.4 Ablation Study
5 Conclusion
References
Nightfall Deception: A Novel Backdoor Attack on Traffic Sign Recognition Models via Low-Light Data Manipulation
1 Introduction
2 Related Work
2.1 Data Poisoning Backdoor Attacks
2.2 Model Poisoning Backdoor Attacks
3 Methodology
3.1 Problem Formulation
3.2 Threat Model
3.3 Our Proposed Backdoor Attack
4 Experiments
4.1 Experimental Setup
4.2 Main Results
4.3 Analysis of the Effect on Network Attention
5 Conclusion
References
Author Index

System requirements

Save as PDF Copy link into clipboard

Schweitzer Fachinformationen

Advanced Data Mining and Applications

Description

More details

Other editions

Additional editions

Content

System requirements