Advances in Knowledge Discovery and Data Mining

Name: Advances in Knowledge Discovery and Data Mining | 23rd Pacific-Asia Conference, PAKDD 2019, Macau, China, April 14-17, 2019, Proceedings, Part III
Brand: Springer
Price: 85.59 EUR
Availability: OnlineOnly

23rd Pacific-Asia Conference, PAKDD 2019, Macau, China, April 14-17, 2019, Proceedings, Part III

Qiang Yang Zhi-Hua Zhou Zhiguo Gong Min-Ling Zhang Sheng-Jun Huang(Herausgeber*in)

Springer (Verlag)

Erschienen am 3. April 2019

XXVIII, 556 Seiten

E-Book

PDF mit Wasserzeichen-DRM

Systemvoraussetzungen

978-3-030-16142-2 (ISBN)

85,59 €inkl. 7% MwSt.

Systemvoraussetzungen

für PDF mit Wasserzeichen-DRM

E-Book Einzellizenz

Als Download verfügbar

Beschreibung

Weitere Details

Weitere Ausgaben

Inhalt

Intro
PC Chairs' Preface
General Chairs' Preface
Organization
Contents - Part III
Representation Learning and Embedding
AAANE: Attention-Based Adversarial Autoencoder for Multi-scale Network Embedding
1 Introduction
2 Preliminaries
3 The Framework
3.1 An Overview of the Framework
3.2 Attention-Based Autoencoder
3.3 Adversarial Learning
3.4 Training Procedure
4 Experiments
4.1 Datasets
4.2 Baselines and Experimental Settings
4.3 Multi-label Classification
4.4 Detailed Analysis of the Proposed Model
5 Related Work
6 Conclusion
References
NEAR: Normalized Network Embedding with Autoencoder for Top-K Item Recommendation
1 Introduction
2 Related Work
2.1 Recommendation Systems
2.2 Network Embedding
3 The Proposed NEAR Model
3.1 Problem Definition
3.2 Framework
3.3 Explicit Relations
3.4 Implicit Relations
3.5 Optimization
3.6 Analysis and Discussion
4 Experiments
4.1 Datasets
4.2 Baseline Algorithms
4.3 Evaluation Metrics
4.4 Experimental Results
5 Conclusion
References
Ranking Network Embedding via Adversarial Learning
1 Introduction
2 Related Work
3 RNE: Ranking Network Embedding
3.1 Framework
3.2 Vanilla Ranking Network Embedding
3.3 Adversarial Ranking Network Embedding
4 Experiments
4.1 Experiment Setup
4.2 Network Visualization
4.3 Link Prediction
4.4 Node Classification
5 Conclusion
References
Selective Training: A Strategy for Fast Backpropagation on Sentence Embeddings
1 Introduction
2 Background
3 Problem Definition
4 Our Approach
5 Evaluation
5.1 Data and Experimental Setup
5.2 Empirical Study and Discussion
6 Detailed Error Analysis
7 Related Work
8 Conclusion and Future Work
References
Extracting Keyphrases from Research Papers Using Word Embeddings
1 Introduction
2 Related Work
3 Proposed Model: WeRank
3.1 Word-Word Graph Embedding
3.2 Word-Topic Graph Embedding
3.3 Topic-Topic Graph Embedding
3.4 Joint Embedding
3.5 Model Optimization
3.6 Word Ranking Using Embedding
4 Experiments
4.1 Experimental Datasets and Settings
4.2 Parameters and Influences
4.3 Results and Analysis
5 Conclusions
References
Sequential Embedding Induced Text Clustering, a Non-parametric Bayesian Approach
1 Introduction
2 Related Work
3 Description of SiDPMM
4 Inference via Collapsed Gibbs Sampling
5 Extraction of Sequential Feature and Synonyms Embedding
6 Experiments
6.1 Empirical Results
7 Conclusion
References
SSNE: Status Signed Network Embedding
1 Introduction
2 Related Work
3 Methodology
3.1 Problem Formulation
3.2 The Model
3.3 Optimization via Negative Sampling
3.4 Training
3.5 Attention-Enhance
3.6 Computational Complexity Analysis
4 Experiments
4.1 Datasets
4.2 Experimental Settings
4.3 Link Sign Prediction
4.4 Effect of Status and Attention
4.5 Parameter Sensitivity
5 Conclusion
References
On the Network Embedding in Sparse Signed Networks
1 Introduction
2 Problem Definition
2.1 Problem Statement
2.2 Challenges
3 sign2vec: A Model for Signed Network Embedding
3.1 Characterizing Similarity in Link Forming Behavior
3.2 Proposed Model
4 Experiments and Evaluation
4.1 Experimental Setup
4.2 Evaluation of Embedding Quality
4.3 Evaluation on Downstream Tasks
5 Conclusion
References
MSNE: A Novel Markov Chain Sampling Strategy for Network Embedding
1 Introduction
2 Preliminary
3 MSNE: A Novel Markov Chain Sampling Strategy for Network Embedding
3.1 n-Order Markov Chain Random Walk
3.2 The MSNE Algorithm
4 Experiments
4.1 Experimental Setup
4.2 Node Classification
4.3 Link Prediction
5 Conclusions
References
Auto-encoder Based Co-training Multi-view Representation Learning
1 Introduction
2 Framework
2.1 Notations
2.2 Core Concept of Framework
2.3 Description of Framework
2.4 Tricks
3 Experiment
3.1 Data Set Partition
3.2 Convergence Analysis of Training Process
3.3 Performance Test on Multiple Learning Task
4 Conclusion
References
Robust Semi-supervised Representation Learning for Graph-Structured Data
1 Introduction
2 Related Work
3 Preliminaries
3.1 Notations
3.2 Graph Convolutional Networks
4 Our Proposed Method
4.1 Enhance Labeled Data
4.2 Large-Margin Cross-Entropy Loss
5 Experiments
5.1 Experimental Setup
5.2 Compared Results
5.3 The Confidence of Assigning Pseudo-Label
5.4 Loss vs. Epoch
6 Discussion
7 Conclusion
References
Characterizing the SOM Feature Detectors Under Various Input Conditions
1 Introduction
2 Related Works
3 Test Setup
3.1 Architecture
3.2 Training and Testing
4 Results and Analysis
4.1 Performance in Constrained Input Environment
4.2 Performance in Perturbed Input Patterns
4.3 Performance in Complex Vehicle Dataset
5 Discussion
References
PCANE: Preserving Context Attributes for Network Embedding
1 Introduction
2 Related Works and Our Motivation
2.1 Related Work
2.2 Our Motivation
3 Problem Definition
4 Proposed Method
4.1 The PCANE Model
4.2 PCANE++: Encoding the Source Attributes Explicitly
4.3 Optimization
5 Experiments
5.1 Experiment Setup
5.2 Link Prediction
5.3 Node Classification
5.4 Discussion
5.5 Parameter Sensitivity
6 Conclusion
References
A Novel Framework for Node/Edge Attributed Graph Embedding
1 Introduction
2 Related Work
3 Problem Formulation
4 Method
4.1 GERI
4.2 SGERI
5 Experiments
5.1 Dataset
5.2 Comparison Algorithm
5.3 Performance of GERI
5.4 Performance of SGERI
5.5 Parameter Analysis
6 Conclusion
References
Mining Unstructured and Semi-structured Data
Context-Aware Dual-Attention Network for Natural Language Inference
1 Introduction
2 Related Work
3 Problem Statement and Model Structure
3.1 Problem Statement
3.2 Context-Aware Dual-Attention Network
3.3 Model Learning
4 Experiments
4.1 Dataset Description
4.2 Overall Performance
4.3 Ablation Performance
4.4 Qualitative Evaluation
5 Conclusion and Future Work
References
Best from Top k Versus Top 1: Improving Distant Supervision Relation Extraction with Deep Reinforcement Learning
1 Introduction
2 Related Work
3 Framework
4 DQN Parameter Learning
5 Experimental Setup and Results
5.1 Dataset
5.2 Baseline Extractors
5.3 RL Models
5.4 Evaluation Metrics
5.5 Experimental Results
5.6 Analysis and Case Study
6 Conclusions
References
Towards One Reusable Model for Various Software Defect Mining Tasks
1 Introduction
2 The Proposed Approach
2.1 Auxiliary Model
2.2 Reusable Model
3 Experiment
3.1 How Good Is RUM
3.2 Reusing RUM for Clone Detection
3.3 Reusing RUM for Defect Prediction
3.4 Reusing RUM for Bug Localization
4 Conclusion
References
User Preference-Aware Review Generation
1 Introduction
2 Related Work
3 Methodology
3.1 Background: Recurrent Neural Networks
3.2 User Preference-Aware Review Generation
4 Experiments
4.1 Experimental Setup
4.2 Results
4.3 Analysis of User Preference
4.4 Case Study
5 Conclusion and Future Work
References
Mining Cluster Patterns in XML Corpora via Latent Topic Models of Content and Structure
1 Introduction
2 Preliminaries
2.1 Traditional Tree-Based XML Document Representation
2.2 XML Features for Topic Modeling
3 MUESLI: A Topic Model for Clustering XML Corpora
3.1 Observed-Data Likelihood and Prior Distributions
3.2 Approximate Posterior Inference and Parameter Estimation
3.3 Partitioning Algorithms
4 PAELLA: Joint XML Clustering and Topic Modeling
5 Evaluation
5.1 XML Corpora, Competitors and Evaluation Measures
5.2 Partitioning Effectiveness
6 Conclusions and Future Research
References
A Large-Scale Repository of Deterministic Regular Expression Patterns and Its Applications
1 Introduction
2 Preliminaries
3 The Data Set
3.1 The Data Sources
3.2 Harvesting Schema Files from the Web
3.3 A Practical Study of DREs
4 The Repository
4.1 Getting the DRE Set from the Data Set
4.2 From the DRE Set to the DRE Pattern Set
4.3 Dynamically Increasing the Power of the Repository
4.4 Getting the Repository of DREs
5 Experiments
5.1 Algorithm Selection
5.2 Experiment1
5.3 Experiment2
6 Conclusion
References
Determining the Impact of Missing Values on Blocking in Record Linkage
1 Introduction
1.1 Our Contributions
2 Related Work
3 Background
4 Major Blocking Methods
5 Experimental Evaluation
5.1 Data Set Construction
5.2 Experimental Setup
5.3 Experimental Results
6 Conclusion
References
Behavioral Data Mining
Bridging the Gap Between Research and Production with CODE
1 Introduction
2 Related Works
3 Background of Online Advertising
4 The CODE Framework
4.1 Communication
4.2 Objectives
4.3 Deliverables
4.4 Evaluations
5 Conclusions and Future Work
References
Distance2Pre: Personalized Spatial Preference for Next Point-of-Interest Prediction
1 Introduction
2 Related Work
3 The Distance2Pre Network
3.1 Problem Formulation
3.2 Sequential Preference
3.3 Spatial Preference
3.4 Preference Encoders
3.5 Training Framework
4 Experiments
4.1 Experimental Settings
4.2 Performance Comparison
4.3 Settings of Max Distance MD and Distance Interval d
4.4 Linear Fusion Vs. Non-Linear Fusion
4.5 Visualization of Spatial Preference
5 Conclusion
References
Using Multi-objective Optimization to Solve the Long Tail Problem in Recommender System
1 Introduction
2 Similarity Measure
3 The Proposed Algorithm
3.1 Objective Function
3.2 Design of the Proposed Algorithm
4 Experiments
4.1 Experimental Settings
4.2 Experimental Results
5 Conclusion
References
Event2Vec: Learning Event Representations Using Spatial-Temporal Information for Recommendation
1 Introduction
2 Problem Definition
3 The Proposed Approach
3.1 Our Models
4 Experiments
4.1 Experimental Setup
4.2 Results
4.3 Impact of Different Factors
4.4 Exploring Various Temporal Patterns
5 Related Work
6 Conclusion
References
Maximizing Gain over Flexible Attributes in Peer to Peer Marketplaces
1 Introduction
2 Preliminaries
2.1 General Problem Definition
2.2 Computational Complexity
3 Solution
4 Gain Function Design
4.1 Frequent-Item Based Count (FBC)
4.2 FBC Computation
5 Experimental Evaluation
5.1 Experimental Setup
5.2 Experimental Results
5.3 Case Study
6 Related Work
7 Final Remarks
References
An Attentive Spatio-Temporal Neural Model for Successive Point of Interest Recommendation
1 Introduction
2 Related Works
2.1 POI Recommendation
2.2 Neural Models and Attention Mechanism
3 Data Description and Analysis
3.1 Data Description
3.2 Check-In Data Exploration
4 Proposed Methodology
4.1 Problem Definition
4.2 POI Embedding and Check-In Representation
4.3 Recurrent Neural Networks and LSTMs
4.4 The Proposed ASTEN Model
4.5 Parameter Inference
5 Experimental Results
5.1 Experimental Setup
5.2 Comparison Methods
5.3 POI Recommendation Performance
5.4 ASTEN Performance Analysis
5.5 Varying Dimensions
6 Conclusion
References
Mentor Pattern Identification from Product Usage Logs
1 Introduction
2 Problem Setup and Feature Extraction
3 Mentor Identification Algorithm
4 Experimental Results
4.1 Dataset Details
4.2 Evaluation Methodology
4.3 Baselines for Comparison
4.4 Comparison of Results
5 Related Work
6 Conclusion
References
Visual Data Mining
AggregationNet: Identifying Multiple Changes Based on Convolutional Neural Network in Bitemporal Optical Remote Sensing Images
1 Introduction
2 Proposed Method
2.1 Network Architecture
2.2 One-Hot Label Map
3 Experiment
3.1 Data and Experimental Settings
3.2 Results and Analysis
4 Conclusion
References
Detecting Micro-expression Intensity Changes from Videos Based on Hybrid Deep CNN
1 Introduction
2 Related Work
3 The Proposed Hybrid Framework
3.1 Pre-processing
3.2 Deep CNN Components
3.3 Fusion Mechanism
4 Results and Analysis
4.1 Micro-expression Intensity Change Detection
4.2 Comparison with Existing Works
5 Conclusion
References
A Multi-scale Recalibrated Approach for 3D Human Pose Estimation
Abstract
1 Introduction
2 Related Work
2.1 3D Human Pose Estimation
2.2 Spatial Enhancement for Deep Networks
3 The Proposed Approach
3.1 Baseline Pose Estimation Network
3.2 Multi-scale Recalibration Regression
3.3 Training
4 Evaluation
4.1 Datasets
4.2 Quantitative Results
4.3 Ablation Study and Analysis
4.4 Qualitative Results
5 Conclusions and Future Work
Acknowledgments
References
Gossiping the Videos: An Embedding-Based Generative Adversarial Framework for Time-Sync Comments Generation
1 Introduction
2 Problem Definition and Technical Solution
2.1 The Embedding Part
2.2 The Generation Part
2.3 Learning the Model
3 Experiments
3.1 Data Preparation
3.2 Experimental Setup
3.3 Overall Results
3.4 Balance for Embedding and Decoding Capacity
3.5 Case Study
4 Related Work
5 Conclusion
References
Self-paced Robust Deep Face Recognition with Label Noise
1 Introduction
2 Related Work
3 Self-paced Robust Face Recognition
3.1 Label Noise
3.2 SPDL Framework
4 Experiments
4.1 Datasets
4.2 Results and Discussion
5 Conclusions
References
Multi-Constraints-Based Enhanced Class-Specific Dictionary Learning for Image Classification
1 Introduction
2 Multi-Constraints-Based Enhanced Class-Specific Dictionary Learning (MECDL)
2.1 Notation
2.2 MECDL Model
2.3 Optimization of MECDL
3 Experiments
3.1 Databases and Experimental Settings
3.2 Results
4 Conclusion
References
Discovering Senile Dementia from Brain MRI Using Ra-DenseNet
1 Introduction
2 Related Work
3 Methods
4 Experiments
4.1 Data Set
4.2 Experimental Setup
4.3 Training
4.4 Results and Discussions
5 Conclusion
References
Knowledge Graph and Interpretable Data Mining
Granger Causality for Heterogeneous Processes
1 Introduction
2 Related Work
3 Theory
3.1 Granger Causality
3.2 Causal Inference by Penalization
3.3 Adaptive Lasso
3.4 Heterogeneous Granger Causality
4 HGGM Algorithm
5 Experimental Results
5.1 Synthetic Heterogeneous Data Sets
5.2 Real-World Applications
6 Conclusions and Future Work
References
Knowledge Graph Embedding with Order Information of Triplets
1 Introduction
2 Related Work
2.1 Researches on Order Information
2.2 Discriminate Models
2.3 Other Models
3 Our Model
3.1 Motivation
3.2 KGE with Recurrent Discriminate Mechanism (RKGE)
3.3 Training
3.4 Complexity Analysis
4 Experiments
4.1 Data Sets
4.2 Knowledge Graph Completion
4.3 Influence of Negative Samples
5 Conclusion and Future Work
References
Knowledge Graph Rule Mining via Transfer Learning
1 Introduction
2 Preliminaries
2.1 Knowledge Graphs and Rules
2.2 Embeddings
3 An Overview of Our Approach
4 Similarity Measures via Embeddings
4.1 Similarity Between Goal Predicates
4.2 Similarity Between Other Predicates
5 Experiments
5.1 From FB15K to YAGO2s
5.2 From FB15K to FB15K
6 Conclusion and Discussion
References
Knowledge Base Completion by Inference from Both Relational and Literal Facts
1 Introduction
2 Background Knowledge
2.1 RDF and RDF KB
2.2 Path Ranking Algorithm
3 Proposed Approach
3.1 Extract Relational Features
3.2 Extract Literal Features
3.3 Prediction Model Training
4 Experiments
4.1 Experimental Settings
4.2 Experiment Results
5 Related Work
6 Conclusion
References
EMT: A Tail-Oriented Method for Specific Domain Knowledge Graph Completion
1 Introduction
2 Related Work
3 EMT Method
3.1 Embedding for MTKG
3.2 EMT by Vectors Multiplication
3.3 EMT by Sparse Matrix
3.4 Time-Space Complexity Analysis
3.5 Training
4 Experiments
4.1 Data Sets
4.2 Link Prediction
4.3 Triplets Classification
5 Conclusion
References
An Interpretable Neural Model with Interactive Stepwise Influence
1 Introduction
2 Related Work
3 Preliminaries
4 Interpretable Neural Networks with Interactive Stepwise Influence
4.1 The Proposed ISI Framework
4.2 Optimization of ISI
5 Experiments
5.1 Data and Setup
5.2 Interpretation Evaluation
5.3 Prediction Accuracy Evaluation
6 Conclusions and Future Work
References
Multivariate Time Series Early Classification with Interpretability Using Deep Learning and Attention Mechanism
1 Introduction
2 Related Work
2.1 Early Classification on Time Series Data
2.2 Time Series Classification Using Deep Learning
3 Proposed Method
3.1 Problem Definition
3.2 MDDNN Model
3.3 Attention Architecture
3.4 Training Process
4 Experiments
4.1 Datasets
4.2 Experimental Settings
4.3 Comparison Methods and Evaluation Metrics
4.4 Experimental Results
4.5 Interpretation
4.6 ECG Interpretation from Doctor's Perspective
5 Conclusion
References
Author Index

Systemvoraussetzungen

Als PDF speichern Als Link merken