Database Systems for Advanced Applications: 26th International Conference, DASFAA 2021, Taipei, Taiwan, April 11–14, 2021, Proceedings, Part I (Lecture Notes in Computer Science, 12681)

✍ Scribed by Christian S. Jensen (editor), Ee-Peng Lim (editor), De-Nian Yang (editor), Wang-Chien Lee (editor), Vincent S. Tseng (editor), Vana Kalogeraki (editor), Jen-Wei Huang (editor), Chih-Ya Shen (editor)

• Publisher: Springer
• Year: 2021
• Tongue: English
• Leaves: 714
• Category: Library

✦ Synopsis

The three-volume set LNCS 12681-12683 constitutes the proceedings of the 26th International Conference on Database Systems for Advanced Applications, DASFAA 2021, held in Taipei, Taiwan, in April 2021.

The total of 156 papers presented in this three-volume set was carefully reviewed and selected from 490 submissions.

The topic areas for the selected papers include information retrieval, search and recommendation techniques; RDF, knowledge graphs, semantic web, and knowledge management; and spatial, temporal, sequence, and streaming data management, while the dominant keywords are network, recommendation, graph, learning, and model. These topic areas and keywords shed the light on the direction where the research in DASFAA is moving towards.

Due to the Corona pandemic this event was held virtually.

✦ Table of Contents

Preface
Organization
Contents – Part I
Contents – Part II
Contents – Part III
Big Data
Learning the Implicit Semantic Representation on Graph-Structured Data
1 Introduction
2 Related Works
3 Semantic Graph Convolutional Networks
3.1 Preliminary
3.2 Latent Factor Routing
3.3 Discriminative Semantic Aggregation
3.4 Independence Learning for Mapping Subspaces
3.5 Algorithm Framework
3.6 Time Complexity Analysis and Optimization
4 Experiments
4.1 Experimental Setup
4.2 Semi-Supervised Node Classification
4.3 Multi-label Node Classification
4.4 Node Clustering
4.5 Visualization Analysis and Semantic-Paths Sampling
5 Conclusion
References
Multi-job Merging Framework and Scheduling Optimization for Apache Flink
1 Introduction
2 Background and Related Work
2.1 Flink DAGs
2.2 Flink Slot
2.3 Related Work
3 Framework Structure
3.1 Model
3.2 Advantages
4 Multi-job Merging and Scheduling
4.1 Multi-job Merging
4.2 Multi-job Scheduling
5 Evaluation Results
5.1 Experimental Setup
5.2 Testing of Multi-job Merging
5.3 Testing of Scheduling Optimization
6 Conclusion and Discussion
References
CIC-FL: Enabling Class Imbalance-Aware Clustered Federated Learning over Shifted Distributions
1 Introduction
2 Related Work
3 The CIC-FL
3.1 System Overview
3.2 Requirements of the Feature for Clustering
3.3 Locally Estimated Global Label Distribution ( LEGLD )
3.4 Bipartition
4 Experiments
4.1 Experimental Settings and Evaluation Metrics
4.2 Experimental Results of CIC-FL
4.3 CIC-FL at Different Levels of Class Imbalance
4.4 CIC-FL at Different Levels of Concept Shift
5 Conclusion
References
vRaft: Accelerating the Distributed Consensus Under Virtualized Environments
1 Introduction
2 Background and Motivation
2.1 The Raft Protocol for Distributed Consensus
2.2 Motivation
3 Design of vRaft
3.1 Overview
3.2 Algorithm Design
3.3 Linearizability of vRaft
4 Implementation and Evaluation
4.1 Experimental Setup
4.2 Overall Results
4.3 Impacts of the Number of Replica
4.4 Impacts of System Load
4.5 Scalability Evaluation
5 Related Work
6 Conclusion
References
Secure and Efficient Certificateless Provable Data Possession for Cloud-Based Data Management Systems
1 Introduction
1.1 Related Work
1.2 Contribution
2 Background
2.1 Elliptic Curve Cryptosystem (ECC)
2.2 Network Model
3 Proposed Scheme
3.1 Setup Algorithm
3.2 Key Generation Algorithm
3.3 Store Algorithm
3.4 Challenge Algorithm
3.5 Generate-Proof Algorithm
3.6 Verify-Proof Algorithm
4 Security Analysis
4.1 Security Model
4.2 Security Theorem
4.3 Discussion
5 Performance Evaluation
5.1 Computation Cost
5.2 Communication Cost
6 Conclusion
References
Dirty-Data Impacts on Regression Models: An Experimental Evaluation
1 Introduction
2 Generalized Evaluation Framework
3 Evaluation Results and Analyses
3.1 Data Sets, Models, and Setup
3.2 Varying Missing Rate
3.3 Varying Inconsistent Rate
3.4 Varying Conflicting Rate
3.5 Lessons Learned
4 Conclusion
References
UniTest: A Universal Testing Framework for Database Management Systems
1 Introduction
2 Related Work
2.1 Benchmark Testing
2.2 Load Testing
3 The UniTest
3.1 User Interface
3.2 Test Management Module
3.3 Test Execution Module
4 Experiments
4.1 Functional Testing
4.2 Performance Testing
5 Conclusion
References
Towards Generating HiFi Databases
1 Introduction
1.1 Hydra
1.2 HF-Hydra
2 LP Formulation
3 Data Generation
4 Experimental Evaluation
4.1 Volumetric Similarity
4.2 Database Summary Overheads
5 Conclusions
References
Modelling Entity Integrity for Semi-structured Big Data
1 Introduction
2 The Running Example
3 Related Work
4 Possibilistic SQL Tables
5 Possibilistic SQL Constraints
6 Conclusion and Future Work
References
Graph Data
Label Contrastive Coding Based Graph Neural Network for Graph Classification
1 Introduction
2 Related Works
3 Proposed Method
3.1 Preliminaries
3.2 LCGNN Architecture Overview
3.3 Label Contrastive Coding
3.4 Graph Encoder Design
3.5 LCGNN Learning
4 Experiments
4.1 Experiment Settings
4.2 Experimental Results and Analysis
5 Conclusion
References
Which Node Pair and What Status? Asking Expert for Better Network Embedding
1 Introduction
2 Related Work
3 Problem Definition
4 Our Solution: ALNE
4.1 ALNE Framework
4.2 Network Embedding: AGCN
4.3 AL Query Strategy
4.4 Information Evaluating
5 Experiments
5.1 Experimental Settings
5.2 Node Classification
5.3 Link Prediction
5.4 Ablation Study
5.5 Parameter Sensitivity
6 Conclusion
References
Keyword-Centric Community Search over Large Heterogeneous Information Networks
1 Introduction
2 Problem Definition
3 Search Algorithm
3.1 The Basic Algorithm
3.2 Advanced Algorithm
3.3 Optimization for the Approaches
4 Experiments
4.1 Experimental Setup
4.2 Effectiveness Testing
4.3 Efficiency Testing
5 Related Work
6 Conclusion
References
KGSynNet: A Novel Entity Synonyms Discovery Framework with Knowledge Graph
1 Introduction
2 Related Work
3 Methodology
3.1 Task Definition
3.2 Semantic Encoder
3.3 Knowledge Encoder
3.4 Fusion Gate
3.5 Similarity Matching and Classification
4 Experiments
4.1 Datasets
4.2 Compared Methods
4.3 Experimental Setup and Evaluation Metrics
4.4 Experimental Results
4.5 Ablation Study
4.6 Online Evaluation
4.7 Case Studies
4.8 Error Analysis
5 Conclusion
References
Iterative Reasoning over Knowledge Graph
1 Introduction
2 Related Works
3 Framework Overview
4 Approach
4.1 Knowledge Graph Constructor
4.2 Iterative Memory Network
4.3 Graph Attention Reasoning
5 Experiment
5.1 Datasets
5.2 Baselines
5.3 Quantitative Study of Commonsense Reason
5.4 Quantitative Study on Question Answering
5.5 Ablation Study
5.6 Case Study
References
Spatial-Temporal Attention Network for Temporal Knowledge Graph Completion
1 Introduction
2 Related Work
2.1 Static Knowledge Graph Embedding Methods
2.2 Temporal Knowledge Graph Embedding Methods
2.3 Deep Spatial-Temporal Models
3 Preliminaries
4 Proposed Model
4.1 Multi-faceted Graph Attention Network
4.2 Adaptive Temporal Attention Mechanism
4.3 Training
5 Experiments
5.1 Experimental Settings
5.2 Performance Comparison
5.3 Model Variants and Ablation Study
5.4 Parameter Analysis
6 Conclusion
References
Ranking Associative Entities in Knowledge Graph by Graphical Modeling of Frequent Patterns
1 Introduction
2 Related Work and Preliminaries
2.1 Related Work
2.2 Definitions and Problem Formulation
3 Methodology
3.1 Structure Learning
3.2 Calculation of Weights
3.3 Ranking Associative Entities
4 Experiments
4.1 Experiment Settings
4.2 Experimental Results
5 Conclusions and Future Work
References
A Novel Embedding Model for Knowledge Graph Completion Based on Multi-Task Learning
1 Introduction
2 Related Work
2.1 Knowledge Graph Completion (KGC)
2.2 Multi-Task Learning
3 Method
3.1 Background and Definition
3.2 Graph Attention Networks (GAT)
3.3 Task-Specific Knowledge Embedding Layer
3.4 Global Shared Layer for Multi-task Learning
3.5 Training
4 Experiments
4.1 Experiment Setup
4.2 Results and Analysis
5 Conclusions
References
Gaussian Metric Learning for Few-Shot Uncertain Knowledge Graph Completion
1 Introduction
2 Related Works
2.1 Completion Methods for DKGs
2.2 Completion Methods for UKGs
2.3 Few-Shot Learning
3 Problem Definition
4 Methodology
4.1 Gaussian Neighbor Encoder
4.2 Gaussian Matching Function
4.3 The Learning Process
5 Experiments
5.1 Datasets
5.2 Baseline Methods
5.3 Experimental Setup
5.4 Link Prediction
5.5 Confidence Prediction
6 Conclusion and Future Work
References
Towards Entity Alignment in the Open World: An Unsupervised Approach
1 Introduction
2 Task Definition and Related Work
3 Methodology
3.1 Side Information
3.2 Unmatchable Entity Prediction
3.3 The Progressive Learning Framework
4 Experiment
4.1 Experiment Settings
4.2 Results
4.3 Ablation Study
4.4 Quantitative Analysis
5 Conclusion
References
Sequence Embedding for Zero or Low Resource Knowledge Graph Completion
1 Introduction
2 Related Work
2.1 Knowledge Graph Embedding
2.2 Pre-trained Language Model
3 Methodology
3.1 Problem Formulation
3.2 Framework Overview
4 Performance Evaluation
4.1 Low Resource Link Prediction
4.2 Open-World KG Completion
5 Conclusions
References
HMNet: Hybrid Matching Network for Few-Shot Link Prediction
1 Introduction
2 Related Work
2.1 Knowledge Graph Embedding
2.2 Few-Shot Learning
3 Problem Definition
4 The Design of HMNet
4.1 Entity-Aware Matching Network
4.2 Relation-Aware Matching Network
4.3 Learning Objective and Algorithm
5 Experiments
5.1 Experimental Setup
5.2 Results
5.3 Further Analysis
5.4 Ablation Study
6 Conclusion
References
OntoCSM: Ontology-Aware Characteristic Set Merging for RDF Type Discovery
1 Introduction
2 Related Work
3 Preliminaries
4 Ontology-Aware Characteristic Set Merging Approach
4.1 Workflow
4.2 Characteristic Set-Based Ontology Extraction
4.3 Ontology-Aware Characteristic Set Merging Algorithm
4.4 Scalability and Complexity
5 Experiments
5.1 Experimental Settings
5.2 Experimental Results
6 Conclusion
References
EDKT: An Extensible Deep Knowledge Tracing Model for Multiple Learning Factors
1 Introduction
2 Related Works
2.1 Single-Factor Models
2.2 Multi-factor Models
3 Preliminaries
3.1 Knowledge Tracing Tasks
3.2 Dynamic Key-Value Memory Networks
4 Model
4.1 Correlation Factors
4.2 Extensible Deep Knowledge Tracing
5 Experiments
5.1 Experimental Setup
5.2 Results and Discussion
6 Conclusion and Future Work
References
Fine-Grained Entity Typing via Label Noise Reduction and Data Augmentation
1 Introduction
2 Problem Definition
3 Framework Overview
4 Graph Based Improvement
4.1 Labeled Training Corpora Noise Reduction
4.2 Labeled Training Corpora Augmentation
5 Sampling Based Improvement
5.1 Labeled Training Corpora Noise Reduction
5.2 Labeled Training Corpora Augmentation
6 Type Prediction via Hierarchical Neural Network
6.1 The Architecture
6.2 Optimization
7 Experiments
7.1 Datasets and Preprocessing
7.2 Comparison and Analysis
8 Related Work
9 Conclusion
References
DMSPool: Dual Multi-Scale Pooling for Graph Representation Learning
1 Introduction
2 Related Work
3 Problem Formulation
4 Methodology
4.1 The Proposed DMSPool Layer
4.2 Multi-scale Graph Convolution Module
4.3 Multi-scale Graph Pooling Module
4.4 Multi-layering
5 Experiment
5.1 Datasets and Baselines
5.2 Parameter Setting
5.3 Performance Comparison on Benchmark Datasets
5.4 Variants of DMSPool
6 Conclusion
References
A Parameter-Free Approach for Lossless Streaming Graph Summarization
1 Introduction
2 Preliminaries
3 Proposed Method
3.1 Framework Overview
3.2 Identify Candidate Supernodes
3.3 Destination Supernode Selection (DSS)
3.4 Update of Summarization
4 Optimizations
4.1 Candidate Supernode Refinement
4.2 Advanced Destination Supernode Selection
5 Experiments
6 Conclusion
References
Expanding Semantic Knowledge for Zero-Shot Graph Embedding
1 Introduction
2 Why RECT Work
2.1 Problem Definition
2.2 Preliminaries: RECT
2.3 RECT-L v.s. ZSL Methods
2.4 The Mechanisms of RECT
3 How to Improve RECT
3.1 The Proposed Method
3.2 Risk Bounds Analysis
4 Experiments
4.1 Setup
4.2 Node Classification
4.3 The Effect of Seen/Unseen Class Number
5 Conclusion
References
Spatial and Temporal Data
Online High-Cardinality Flow Detection over Big Network Data Stream
1 Introduction
2 Problem Statement
3 Design of Online High-Cardinality Flow Detection
3.1 Data Structure
3.2 Algorithm Design
3.3 Online High-Cardinality Flow Detection
4 Optimal System Parameters
4.1 Report Probability
4.2 Constraints for System Parameters
5 Experimental Evaluation
5.1 Experiment Setup
5.2 Comparison in Terms of Memory Requirements
5.3 Comparison in Terms of High-Cardinality Flow Detection
6 Related Work
7 Conclusion
References
SCSG Attention: A Self-centered Star Graph with Attention for Pedestrian Trajectory Prediction
1 Introduction
2 Problem Definitions and Important Notations
3 Methodology
3.1 SCSG Attention Framework
3.2 Spatial and Temporal Encoder
3.3 Attention Mechanism
3.4 Self-centered Star Graph Decoder
4 Experiments and Analysis
4.1 Experimental Setup
4.2 Performance Evaluation
4.3 Case Study
5 Related Work
5.1 RNN Based Sequence Model
5.2 Social Interaction Awareness Model
5.3 Attention Model
6 Conclusion
References
Time Period-Based Top-k Semantic Trajectory Pattern Query
1 Introduction
2 Problem Formulation
3 Baseline Method
4 Our Approach
4.1 Cluster Validity Check
4.2 Algorithm
5 Experimental Evaluation
5.1 Efficiency Study
5.2 Case Study
6 Related Work
7 Conclusion
References
Optimal Sequenced Route Query with POI Preferences
1 Introduction
2 Related Work
2.1 Optimal Route Queries
2.2 Indexes for Road Networks
3 Problem Formalization
4 Baseline for RCOSR
5 Recurrent Optimal Subroute Expansion
5.1 Optimal Subroute Expansion Algorithm
5.2 Reference Node Inverted Index
5.3 The ROSE Algorithm
5.4 Complexity Analysis
6 Experiments
6.1 Efficiency of RCOSR Algorithms
6.2 Efficiency of RNII Index
7 Conclusion
References
Privacy-Preserving Polynomial Evaluation over Spatio-Temporal Data on an Untrusted Cloud Server
1 Introduction
2 Related Work
3 Problem Statement
3.1 System Model
3.2 Threat Model and Design Goals
4 Privacy-Preserving Polynomial Evaluation over Spatio-temporal Data
4.1 Paillier Homomorphic Encryption
4.2 Order-Revealing Encryption
4.3 The Proposed Encryption Scheme
4.4 Virtual Road Network
4.5 Encrypting Spatio-temporal Data over Virtual Road Network
4.6 Executing Polynomials over Encrypted Spatio-temporal Data
5 Security and Performance Analysis
5.1 Security Analysis
5.2 Performance Analysis
6 Experimental Study
6.1 Experimental Setup
6.2 Encrypting Costs
6.3 Time Costs of Executing Polynomial Evaluation
6.4 Overheads of Communication
7 Conclusion
References
Exploiting Multi-source Data for Adversarial Driving Style Representation Learning
1 Introduction
2 Related Work
3 Problem Statement
3.1 Definitions and Notations
3.2 Problem Statement
4 The Design
4.1 Overview
4.2 GPS Data Transformation
4.3 Driving Context Representation
4.4 Learning Model
5 Performance Evaluation
5.1 Experimental Setup
5.2 Driver Number Estimation
5.3 Driver Identification
6 Conclusion
References
MM-CPred: A Multi-task Predictive Model for Continuous-Time Event Sequences with Mixture Learning Losses
1 Introduction
2 Related Work
3 Continuous-Time Event Sequence
4 Methodology
4.1 RNN Encoder and CNN Encoder
4.2 Generators and Discriminator
4.3 Training Strategy
5 Experiments
5.1 Datasets
5.2 Compared Methods
5.3 Metrics
5.4 Experiments on Event Prediction
5.5 Experiments on Time Prediction
5.6 Ablation Study
5.7 Discussion of Sequence Length
5.8 Experimental Details
6 Conclusion and Future Work
References
Modeling Dynamic Social Behaviors with Time-Evolving Graphs for User Behavior Predictions
1 Introduction
2 Related Work
2.1 User Behavior Modeling
2.2 Graph Convolution Network-Based Prediction Models
3 Problem Statement
4 Proposed Model
4.1 Overview
4.2 Modeling User Dynamic Social Behaviors
4.3 Modeling User Similarities in Demographics
4.4 Modeling Individual-Level Behavior Patterns
4.5 Generating Prediction Results
4.6 Model Learning
5 Experiments
5.1 Dataset
5.2 Baselines and Evaluation Metrics
5.3 Experimental Setting
5.4 Experimental Results
6 Conclusion and Future Work
References
Memory-Efficient Storing of Timestamps for Spatio-Temporal Data Management in Columnar In-Memory Databases
1 Introduction
2 Related Work
3 Data Layouts for Timestamps in Columnar Databases
3.1 Common Data Layouts for Timestamps
3.2 A Multiple Column Approach to Store Timestamps
4 Evaluation
4.1 Dataset
4.2 Impact of Different Compression Techniques on the Memory Consumption of the Data Layouts
4.3 Impact of Different Compression Techniques on the Runtime Performance of the Data Layouts
5 Workload-Aware Optimizations of Data Layouts for Timestamps in Columnar In-Memory Databases
5.1 Workload-Driven Combined Data Layout and Compression Scheme Optimization for Timestamps
5.2 Workload-Driven Compression Scheme Selection for Storing Timestamps in a Multiple Column Data Layout
6 Conclusions and Future Work
References
Personalized POI Recommendation: Spatio-Temporal Representation Learning with Social Tie
1 Introduction
2 Related Work
3 Problem Definition
4 Methodology
4.1 Heterogeneous Graph Construction
4.2 Learning Latent Representation
4.3 Modeling User Dynamic and Personalized Preference
4.4 Personalized POI Recommendation
5 Experiments
5.1 Datasets
5.2 Evaluation Metrics
5.3 Baselines
5.4 Parameter Setting
5.5 Performance Comparison
5.6 Ablation Study
5.7 Sensitivity of Hyper-parameters
6 Conclusion
References
Missing POI Check-in Identification Using Generative Adversarial Networks
1 Introduction
2 Related Work
2.1 Spatial Missing Data Imputation
2.2 POI Recommendation
3 Problem Formulation
4 Methodology
4.1 Bi-directional GRU Cell with Time Decay
4.2 The Generator Module
4.3 The Discriminator Module
5 Experiments
5.1 Experimental Setup
5.2 Performance Comparison
5.3 Ablation Analysis
5.4 Parameter Analysis
6 Conclusion
References
Efficiently Discovering Regions of Interest with User-Defined Score Function
1 Introduction
2 Problem Statement
2.1 Preliminaries
2.2 Radius Bounded ROI (RBR) Queries
2.3 Discussion of the Region Score
3 Query Processing Algorithms
3.1 Baseline Algorithm: PairEnum
3.2 Circle Rotation and Angle Scan
3.3 Algorithm: BaseRotation
3.4 Algorithm: OptRotation
4 Experimental Studies
4.1 Experiment Setting
4.2 Experiment Results
5 Conclusion
References
An Attention-Based Bi-GRU for Route Planning and Order Dispatch of Bus-Booking Platform
1 Introduction
2 Related Work
3 Problem Formulation
4 Attention-Based Bi-GRU Method
4.1 Framework Overview
4.2 Order Information Extraction
4.3 Attention Sub-network
4.4 Order Dispatch and Route Planning
5 Experiments and Analysis
5.1 Dataset and Evaluation Criteria
5.2 Destination Stations Selection
5.3 Simulations
6 Conclusion
References
Top-k Closest Pair Queries over Spatial Knowledge Graph
1 Introduction
2 Preliminaries
3 Problem Definition and Basic Algorithm
3.1 Problem Definition
3.2 Basic Solution
4 Improved Solution
4.1 Vertex Join Node Method: V2N
4.2 Improved Vertex Join Node Method: V2N
5 Experiment
5.1 Settings
5.2 Efficiency Evaluation
6 Related Work
7 Conclusion
References
HIFI: Anomaly Detection for Multivariate Time Series with High-order Feature Interactions
1 Introduction
2 Related Work
3 Methodology
3.1 Multivariate Feature Interaction Module
3.2 Attention-Based Time Series Modeling Module
3.3 Variational Encoding Module
3.4 Model Training
4 Experiments
4.1 Experimental Setup
4.2 Overall Performance Comparison
4.3 Ablation Study
5 Conclusion
References
Incentive-aware Task Location in Spatial Crowdsourcing
1 Introduction
2 Problem Statement
3 Proposed Methods
3.1 Even Clustering Location Method
3.2 Uneven Clustering Location Method
3.3 Uneven Greedy Location Method
4 Experimental Study
4.1 Experimental Methodology
4.2 Experiments on Real Data
5 Conclusion
References
Efficient Trajectory Contact Query Processing
1 Introduction
2 Related Works
3 Problem Statement
4 Iteration-Based Trajectory Contact Search
5 Advanced Contact Search Algorithm
5.1 Hop Scanning Algorithm
5.2 Time Interval Grid Index
6 Experiments
6.1 Experiment Settings
6.2 Effectiveness Study
7 Conclusion
References
STMG: Spatial-Temporal Mobility Graph for Location Prediction
1 Introduction
2 Related Work
3 Proposed Model
3.1 Problem Definition
3.2 Spatial-temporal Mobility Graph Construction
3.3 Spatial-temporal Enhanced Graph Neural Network
4 Experiments
4.1 Datasets and Settings
4.2 Performance Comparison
5 Conclusion
References
Author Index