Computational Science – ICCS 2023: 23rd International Conference, Prague, Czech Republic, July 3–5, 2023, Proceedings, Part I (Lecture Notes in Computer Science)

Preface
Organization
Contents – Part I
ICCS 2023 Main Track Full Papers
Improving the Resiliency of Decentralized Crowdsourced Blockchain Oracles
	1 Introduction
	2 Related Work
	3 Proposed Model
		3.1 System Overview
		3.2 Agents
		3.3 Reputation
		3.4 Threat Models
		3.5 Rewards
		3.6 Evaluation
	4 Experiments and Simulation
		4.1 Simulation Settings
		4.2 Participation Control
		4.3 Weighted Voting
		4.4 Stratified Voting
	5 Discussion
	6 Conclusion
	References
Characterization of Pedestrian Contact Interaction Trajectories
	1 Introduction
	2 Datasets
	3 Data Analysis
	4 Conclusion
	References
Siamese Autoencoder-Based Approach for Missing Data Imputation
	1 Introduction
	2 Related Work
	3 Siamese Autoencoder-Based Approach for Imputation
		3.1 Deep Autoencoder Architecture
		3.2 Custom Loss Function
		3.3 Custom Triplet Mining
	4 Experimental Setup
	5 Results
	6 Conclusions
	References
An Intelligent Transportation System for Tsunamis Combining CEP, CPN and Fuzzy Logic
	1 Introduction
	2 Related Work
	3 Application Scenario
	4 The CEP Event Patterns
	5 The Fuzzy Inference System
	6 CPN Model
	7 Conclusions and Future Work
	References
Downscaling WRF-Chem: Analyzing Urban Air Quality in Barcelona City
	1 Introduction
	2 Data, Materials and Methods
		2.1 Case Study
		2.2 Model Description, Chemistry and Physics Schemes
	3 Experimental Results
		3.1 Meteorology Results
		3.2 Air Quality Results
	4 Conclusions
	References
Influence of Activation Functions on the Convergence of Physics-Informed Neural Networks for 1D Wave Equation
	1 Introduction
	2 Wave Equation
	3 Training
	4 Numerical Results
	5 Experiments
		5.1 Parameters Tuning
		5.2 Activation Functions
	6 Results
	7 Conclusions and Future Work
	References
Accelerating Multivariate Functional Approximation Computation with Domain Decomposition Techniques
	1 Introduction
		1.1 Related Work
	2 Approach
		2.1 Numerical Background
		2.2 Shared Knot Spans at Subdomain Interfaces
		2.3 Solver Workflow
		2.4 Implementation
	3 Results
		3.1 Error Convergence Analysis
		3.2 Real Simulation Datasets
		3.3 Parallel Scalability
	4 Summary
	References
User Popularity Preference Aware Sequential Recommendation
	1 Introduction
	2 Related Works
		2.1 Sequential Recommendation
		2.2 Popularity Aware Recommendation
		2.3 Contrastive Learning
	3 Proposed Method
		3.1 Problem Statement
		3.2 Basic Model
		3.3 Sequential Popularity Perception Module
		3.4 Popularity Contrastive Learning Module
		3.5 Network Training
	4 Experiment
		4.1 Datasets
		4.2 Baselines
		4.3 Implementation Details and Evaluation Metrics
		4.4 Performance Comparison
		4.5 Performance on Particular Users
		4.6 Ablation Study
	5 Conclusion
	References
Data Heterogeneity Differential Privacy: From Theory to Algorithm
	1 Introduction
	2 Related Work
	3 Preliminaries
		3.1 Notations and Assumptions
		3.2 Differential Privacy
	4 Sharper Utility Bounds for DP-SGD
	5 Performance Improving DP-SGD
		5.1 Influence Function and Error Analysis
		5.2 Performance Improving DP-SGD
		5.3 Privacy Guarantees
		5.4 Utility Analysis
	6 Comparison with Related Work
	7 Experimental Results
	8 Conclusions
	References
Inference of Over-Constrained NFA of Size k+1 to Efficiently and Systematically Derive NFA of Size k for Grammar Learning
	1 Introduction
	2 The NFA Inference Problem
		2.1 Notations
		2.2 A ``Meta-model''
		2.3 Some Previous Models
	3 k_NFA Extensions
		3.1 Building a (k+1)_NFA from a k_NFA
		3.2 (k+1)_NFA+ Extension
		3.3 k_NFA Extension
		3.4 Complexity
	4 Properties of the Extensions
		4.1 (k+1)_NFA+
		4.2 (k+1)_NFA
	5 Experimentation
		5.1 Context for Reproductibility
		5.2 Simplified Models
		5.3 Results and Discussions
	6 Conclusion
	References
Differential Dataset Cartography: Explainable Artificial Intelligence in Comparative Personalized Sentiment Analysis
	1 Introduction
	2 Background
		2.1 Personalization in NLP
		2.2 Explainable AI
	3 Datasets
	4 Personalized Architectures
	5 HumAnn
	6 Differential Data Maps
	7 Experimental Setup
	8 Results
	9 Conclusions and Future Work
	References
Alternative Platforms and Privacy Paradox: A System Dynamics Analysis
	1 Introduction
	2 Theoretical Background
		2.1 Privacy as a Social Issue
		2.2 Social Theory Based Explanations of the Privacy Paradox
	3 A System Dynamics Model of the Privacy Paradox
		3.1 Problem Articulation and Dynamic Hypothesis
	4 Model Development
		4.1 Model Structure
		4.2 Model Parameters
		4.3 Model Testing and Validation
	5 Simulation Results
		5.1 Simulation Experiment 1
		5.2 Simulation Experiment 2
		5.3 Simulation Experiment 3
	6 Concluding Discussion
	References
The First Scientiffic Evidence for the Hail Cannon
	1 Introduction
	2 Experimental Verification
	3 Numerical Simulations
	4 IGA-ADS Simulation of the Hail Cannon
	5 Conclusion and Future Work
	References
Constituency Parsing with Spines and Attachments
	1 Introduction
	2 Headed Constituencies
	3 The Dataset
	4 Proposed Parsing Technique
		4.1 Spines
		4.2 Spine Based Parsing
	5 Parser Architecture
	6 Related Work
	7 Evaluation
	8 Conclusions
	References
Performing Aerobatic Maneuver with Imitation Learning
	1 Introduction
	2 Related Work
	3 Data Analysis
		3.1 Maneuvers Description
		3.2 Evaluation Metrics
		3.3 Maneuvers Evaluation
	4 Controllers Training
		4.1 Results
		4.2 Discussion of Results
	5 Circuit Controller
	6 Conclusion
	References
An Application of Evolutionary Algorithms and Machine Learning in Four-Part Harmonization
	1 Introduction
		1.1 State of the Art
		1.2 Contribution
	2 Soprano Harmonization Problem
	3 Algorithmic Approach
		3.1 Genetic Algorithm
		3.2 Bayesian Network
		3.3 Hybrid Algorithm
	4 Test Results
	5 Conclusions
	References
Predicting ABM Results with Covering Arrays and Random Forests
	1 Introduction
	2 Related Work
	3 Approach
		3.1 Heatbugs Model
		3.2 Choosing Parameters via Covering Arrays
		3.3 Machine Learning
	4 Experimental Setup
		4.1 Data Gathering and Preparation
		4.2 Machine Learning in All Experiments
	5 Results
		5.1 Experiment A: Low Unhappiness and Low Variation
		5.2 Experiment B: Steady Unhappiness
		5.3 Experiment C: Average Unhappiness
		5.4 Feature Importance
	6 Conclusions
	References
Vecpar – A Framework for Portability and Parallelization
	1 Introduction
	2 State of the Art and Related Work
	3 Proposed Approach
	4 Evaluation
		4.1 BabelStream Benchmark
		4.2 Vecpar Internal Benchmark
		4.3 Track Reconstruction Use Cases
	5 Conclusions and Future Work
	References
Self-supervised Deep Heterogeneous Graph Neural Networks with Contrastive Learning
	1 Introduction
	2 Related Work
		2.1 Heterogeneous Graph Neural Networks
		2.2 Contrastive Learning
	3 Preliminary
	4 The Proposed DHG-CL Model
		4.1 Node Transformation
		4.2 Cross-Layer Semantic Encoder
		4.3 Graph-Based Contrastive Learning
	5 Experiments
		5.1 Experimental Setup
		5.2 Node Classification
		5.3 Node Clustering
		5.4 Visualization
		5.5 Variant Analysis
		5.6 Parameter Analysis
	6 Conclusion and Future Work
	References
First-Principles Calculation to N-type Beryllium Related Co-doping and Beryllium Doping in Diamond
	1 Introduction
	2 Calculation Methods
	3 Results and Discussion
		3.1 Impurity Formation Energy (Ef)
		3.2 Ionization Energies
		3.3 Electronic Structure
		3.4 Band Structure
	4 Conclusions
	References
Machine Learning Detects Anomalies in OPS-SAT Telemetry
	1 Introduction
	2 Materials and Methods
		2.1 Dataset
		2.2 Detecting OPS-SAT Anomalies Using Machine Learning
	3 Experimental Validation
		3.1 Experiment 1: Exploiting Original Training Dataset
		3.2 Experiment 2: Augmenting Training Datasets
	4 Conclusions and Future Work
	References
Wildfire Perimeter Detection via Iterative Trimming Method
	1 Introduction
	2 Thermal Infrared Image of a Wildfire
	3 Delaunay Triangulation and Iterative Trimming
		3.1 Delaunay Triangulation
		3.2 Iterative Trimming Method
	4 Results and Discussion
		4.1 Iterative Trimming Method
		4.2 Canny Edge Detector
		4.3 Graph-Cut Method
		4.4 Level Set Method
	5 Conclusions
	References
Variable Discovery with Large Language Models for Metamorphic Testing of Scientific Software
	1 Introduction
	2 State of the Art
	3 Discovering I/O Variables with an LLM
		3.1 LLM-Based Workflow
		3.2 Prompt Construction and LLM Particulars
	4 Evaluation
		4.1 Ground Truth and Experiment Setup
		4.2 Results
		4.3 Discussion and Threats to Validity
	5 Conclusion and Future Work
	References
On Irregularity Localization for Scientific Data Analysis Workflows
	1 Introduction
	2 Motivation and Background
		2.1 General Framework for Outcome-Preserving Input Reduction
	3 Instantiation of the General Framework
	4 Investigated Reduction Strategies
		4.1 Baseline (Leave-One-Out)
		4.2 Delta Debugging (dd-min)
		4.3 Probabilistic Delta Debugging (prob-dd)
		4.4 Similarity-Based Isolation (similarity-iso)
	5 Evaluation
		5.1 Experimental Setup
		5.2 Results
		5.3 Summary and Threats to Validity
	6 Related Work
	7 Conclusion and Future Work
	References
RAFEN – Regularized Alignment Framework for Embeddings of Nodes
	1 Introduction
	2 Related Work
	3 The Proposed RAFEN Framework
	4 Experimental Setup
		4.1 Datasets
		4.2 Node Embeddings
		4.3 Aligned Models
		4.4 Link Prediction
	5 Results
		5.1 Embeddings Aggregation Method Comparison
		5.2 Comparison of RAFEN's -Based Variants
		5.3 RAFEN Comparison to Baselines
	6 Conclusion and Future Work
	References
CLARIN-Emo: Training Emotion Recognition Models Using Human Annotation and ChatGPT
	1 Introduction
	2 Related Work
	3 Datasets
		3.1 CLARIN-Emo: Human Texts and Annotations
		3.2 Stockbrief-GPT: Human Texts Annotated by ChatGPT
		3.3 ChatGPT-Emo: ChatGPT Texts and Annotations
	4 Models
		4.1 Sequential Sentence Classification
		4.2 Text Classification
	5 Experiments
		5.1 Sequential Sentence Classification
		5.2 Text Classification
	6 Results
	7 Conclusions and Future Work
	References
B2-FedGAN: Balanced Bi-directional Federated GAN
	1 Introduction
	2 Related Work
	3 Bi-directional FedAvg Based Balanced GAN for Learning Process
	4 Experimental Setup
		4.1 Data Collection
		4.2 Results
	5 Conclusion
	References
Graph-Level Representations Using Ensemble-Based Readout Functions
	1 Introduction
	2 Related Work
		2.1 Graph Neural Networks
		2.2 Graph-Level Prediction
	3 Graph Neural Network Readouts
		3.1 Problem Statement
		3.2 Non-parametrized Readout Functions
		3.3 Parametrized Readout Functions
		3.4 Ensemble Readouts
	4 Experiments
		4.1 Datasets
		4.2 Evaluation Protocol
		4.3 Results
	5 Conclusion
	References
Parallel Adjoint Taping Using MPI
	1 Introduction
	2 Notation and Foundations
	3 Serial Adjoint Reversal of Evolution
		3.1 Reversal of Serial Evolution
	4 Parallel Adjoints of Evolution
		4.1 Parallel Reversal Procedure
		4.2 Run Time Analysis
	5 Case Study: Lorenz Attractor
	6 Case Study: Parallel Taping in OpenFOAM
	7 Summary and Outlook
	References
A Moral Foundations Dictionary for the European Portuguese Language: The Case of Portuguese Parliamentary Debates
	1 Introduction
	2 Related Work
	3 European Portuguese Moral Foundations Dictionary
		3.1 Semi-supervised Translation of the English MFD
		3.2 Experimental Validation
	4 Morality in the Portuguese Assembly of the Republic, a Case Study
		4.1 Parliamentary Transcripts
		4.2 Measuring Moral Load
		4.3 Political Analysis
	5 Conclusions
	References
Towards Automatic Generation of Digital Twins: Graph-Based Integration of Smart City Datasets
	1 Introduction
	2 Motivation
	3 Formal Background
		3.1 Spatially-Triggered Graph Transformations
		3.2 Notations
	4 Solution Outline
		4.1 OpenStreetMap Graph Generation
		4.2 Sensor Modelling
		4.3 Sensor Assignment
		4.4 Traffic Flow Modelling
	5 Conclusions and Future Work
	References
Forecasting Cryptocurrency Prices Using Contextual ES-adRNN with Exogenous Variables
	1 Introduction
	2 Data and Forecasting Problem
	3 Model
		3.1 Main Track
		3.2 Context Track
	4 Experimental Study
	5 Conclusions
	References
Enhanced Emotion and Sentiment Recognition for Empathetic Dialogue System Using Big Data and Deep Learning Methods
	1 Introduction
	2 Terabot – A Therapeutic Dialogue System
		2.1 Sentiment and Emotion Classification
	3 Datasets
		3.1 CORTEX Dataset
		3.2 Common Crawl Dataset
	4 Experiments
		4.1 Improving the Classification Model
		4.2 Dataset Expansion Process
	5 Results and Discussion
		5.1 Impact of Improved Language Model
		5.2 Impact of Extended Training Dataset
	6 Conclusions
	References
Automatic Delta-Adjustment Method Applied to Missing Not At Random Imputation
	1 Introduction
	2 Related Work
	3 Automatic Delta-Adjustment Method
	4 Experimental Setup
	5 Results
	6 Conclusions
	References
Data Integration Landscapes: The Case for Non-optimal Solutions in Network Diffusion Models
	1 Introduction
	2 Background and Related Work
	3 Data Integration Landscapes
		3.1 Entity Resolution
		3.2 Graph Resolution
		3.3 Multiple Resolution Functions
	4 Probability Assignment
	5 Experiment
		5.1 Instantiation
		5.2 Dataset
		5.3 Model
		5.4 Results
	6 Conclusions
	References
Excessive Internet Use in the Organizational Context: A Proposition of the New Instrument to Measure Cyberloafing at Work
	1 Introduction
		1.1 Dimensionality of Cyberloafing
		1.2 Antecedents and Consequences of Cyberloafing
	2 Method
		2.1 Participants and Procedure
		2.2 Measures
	3 Results
		3.1 Factor Structure and Reliability of the CBLS-15 Measure
		3.2 Cyberloafing and Work-Related Characteristics
	4 Discussion
		4.1 Practical Implications
		4.2 Limitations and Future Directions
	5 Conclusions
	References
FAIR-FATE: Fair Federated Learning with Momentum
	1 Introduction
	2 Related Work
	3 Problem Statement
		3.1 Federated Learning and Momentum
		3.2 Fairness Metrics
		3.3 Objective
	4 FAIR-FATE: Fair Federated Learning with Momentum
	5 Experiments
		5.1 Datasets
		5.2 Non-identical Client Data
		5.3 Implementation and Setup Details
		5.4 Experimental Results
	6 Conclusion and Future Work
	References
Multi-agent Cellular Automaton Model for Traffic Flow Considering the Heterogeneity of Human Delay and Accelerations
	1 Introduction
	2 Model
		2.1 Parameters
		2.2 Limiting Distances
		2.3 Dynamics
		2.4 Agents Behaviour
	3 Numerical Results
	4 Conclusion
	References
A Contrastive Self-distillation BERT with Kernel Alignment-Based Inference
	1 Introduction
	2 Related Work
	3 Method
		3.1 Early Exit
		3.2 Classification Loss
		3.3 Contrastive Loss
		3.4 Distillation Loss
		3.5 Total Loss
		3.6 Centered Kernel Alignment
	4 Experiments
		4.1 Experimental Results on GLUE and ELUE
		4.2 Ablation Study
		4.3 Parameter Analysis
	5 Conclusion
	References
Turning Flight Simulation with Fluid-Rigid Body Interaction for Flying Car with Contra-Rotating Propellers
	1 Introduction
	2 Numerical Approach
		2.1 Governing Equations
		2.2 Moving Computational Domain Approach
		2.3 Coupled Simulation with Rigid Body
	3 Flight Simulation of eVTOL Flying Car
		3.1 Computational Model
		3.2 Control Model
		3.3 Flight Condition
	4 Result and Discussion
	5 Conclusions
	References
Combining Outlierness Scores and Feature Extraction Techniques for Improvement of OoD and Adversarial Attacks Detection in DNNs
	1 Introduction
	2 Related Work
	3 Method
		3.1 Different Categories of OoD Detection Methods
		3.2 Feature Extraction
		3.3 Proposed Ensemble OoD Detector
	4 Experiments and Results
		4.1 OoD Detection Problem
		4.2 Attacks
		4.3 Sensitivity Analysis
		4.4 Limitations
	5 Conclusion
	References
From Online Behaviours to Images: A Novel Approach to Social Bot Detection
	1 Introduction
	2 Related Work
	3 Useful Notions
		3.1 Digital DNA
		3.2 Convolutional Neural Networks
	4 From Digital DNA to Images
	5 Datasets
		5.1 Cresci-2017
		5.2 Cresci-Stock 2018
		5.3 TwiBot20
	6 Experiments and Results
	7 Conclusions
	References
Linking Scholarly Datasets—The EOSC Perspective
	1 Introduction
	2 The EOSC and Open Scholarly Datasets
		2.1 The EOSC and the EOSC Future Project
		2.2 Scholarly Datasets
	3 Related Work
	4 The OARGLink Framework—Problem Statement and General Idea
	5 Data Processing Flow in the OARGLink
	6 The OARGLink Customization
		6.1 Linking Samples of the OARG and AminerNetwork
		6.2 Linking Samples of the OpenAIRE and MAG
	7 The Overall Results and Their Verification
	8 Discussion
	9 Summary and Future Work
	References
Memory-Efficient All-Pair Suffix-Prefix Overlaps on GPU
	1 Introduction
	2 Background and Related Work
	3 Methodology
	4 GPU Implementation
		4.1 Overview
		4.2 Sorting and Searching in Hybrid-Memory
	5 Results
		5.1 Datasets and Testbed
		5.2 Execution Times
		5.3 Scalability
		5.4 Comparison of Execution Times with Other Tools
	6 Conclusions
	References
Ensemble Based Learning for Automated Safety Labeling of Prescribed Fires
	1 Introduction
	2 QUIC-Fire Output Data
	3 Feature Definitions
	4 Automatic Labeling Algorithm
		4.1 Postprocessing of QUIC-Fire Output
		4.2 Process of Automatic Labeling
	5 Optimization
	6 Numerical Results
		6.1 Ensemble Based Learning
		6.2 Automated Safety Labeling
		6.3 Re-evaluation of Manual Labeling
		6.4 Further Improvements
	7 Conclusions
	References
SLAM Methods for Augmented Reality Systems for Flight Simulators
	1 Introduction
	2 Materials and Methods
	3 Results and Discussion
	4 Conclusion
	References
Improving the Performance of Task-Based Linear Algebra Software with Autotuning Techniques on Heterogeneous Architectures
	1 Introduction
	2 The Chameleon Library
	3 Self-optimization Methodology
		3.1 Selecting the Search Strategy
		3.2 Training the Routines
		3.3 Validating the Methodology
	4 Experimental Study
		4.1 Selecting the Routine Parameters
		4.2 Selecting the System and Scheduling Parameters
		4.3 Validating the Methodology
	5 Conclusions
	References
Tempo and Time Signature Detection of a Musical Piece
	1 Introduction
		1.1 Related Work
		1.2 Contribution
	2 Tempo and Time Signature Detection Method
	3 Quantitative Experiments
		3.1 Dataset
		3.2 Effectiveness Evaluation Criteria
		3.3 Tempo Detection
		3.4 Time Signature Detection
	4 Conclusions and Future Work
	References
Correction to: Alternative Platforms and Privacy Paradox: A System Dynamics Analysis
	Correction to: Chapter “Alternative Platforms and Privacy Paradox: A System Dynamics Analysis” in J. Mikyška et al. (Eds.): Computational Science – ICCS 2023, LNCS 14073, https://doi.org/10.1007/978-3-031-35995-8_12
Author Index