Recent Challenges in Intelligent Information and Database Systems: 13th Asian Conference, ACIIDS 2021, Phuket, Thailand, April 7–10, 2021, Proceedings

Preface
Conference Organization
Contents
Data Mining and Machine Learning Methods
CNN Based Analysis of the Luria's Alternating Series Test for Parkinson's Disease Diagnostics
	1 Introduction
	2 Background
		2.1 Luria's Alternating Series Tests
	3 Problem Statement
	4 Experimental Setting
		4.1 Tested Subjects
		4.2 Data Acquisition
	5 Proposed Workflow and Methods
		5.1 Enhancement and Augmentation
		5.2 Workflow of CNN Training Testing and Validation
	6 Main Results
	7 Discussion
	8 Conclusions
	References
Examination of Water Temperature Accuracy Improvement Method for Forecast
	1 Introduction
	2 Background
		2.1 Importance of Water Temperature
		2.2 Proposed System
		2.3 Water Temperature Observation Device
		2.4 Seawater Temperature Announcement System
		2.5 Purpose
	3 Seawater Temperature Prediction
		3.1 Issues
		3.2 Detection and Removement of Outliers
		3.3 Improvement of Interpolation Accuracy
	4 Evaluation of Proposed Method 1
		4.1 Evaluation Method
	5 Future Works
		5.1 Outlier Detection and Removement Method
		5.2 Improvement of Interpolation Accuracy
	References
Assembly Process Modeling Through Long Short-Term Memory
	1 Introduction
	2 Related Work
	3 Next Assembly Step Prediction Through LSTM
	4 Experimental Results
	5 Conclusions and Further Work
	References
CampusMaps: Voice Assisted Navigation Using Alexa Skills
	1 Introduction
	2 Background and Terminology
		2.1 Voice Assistants
		2.2 Navigation Systems
		2.3 Terminology
	3 Proposed System: CampusMaps
		3.1 Architecture
		3.2 Modelling Maps
		3.3 Routing
		3.4 Alexa Skill Development
	4 Advantages and Challenges
		4.1 Advantages
		4.2 Challenges
	5 Results
		5.1 Alpha Tests
		5.2 Beta Tests
	6 Conclusion and Future Scope
	References
Advanced Data Mining Techniques and Applications
Solving Reduction Problems in Cover Lattice Based Decision Tables
	1 Introduction
	2 Covering Based Decision Tables
		2.1 Covering Based Rough Sets and Covering Decision Systems
		2.2 Cover Lattice Based Decision Table
		2.3 An Example: Cover Lattice Based Decision Tables in Transaction Databases
	3 The Reduction of the Condition Lattice and the Fitting of the Decision Lattice
		3.1 Reduction in Cover Lattice Based Decision Tables
		3.2 Decision Fitting in the Cover Lattice Based Decision Table
		3.3 A Typical Application of Cover Lattice Based Rough Set
	4 Related Work
	5 Conclusions and Future Work
	References
Forecasting System for Solar-Power Generation
	1 Introduction
	2 Literature Review
		2.1 Solar-Power Generation and Prediction
		2.2 Machine Learning Algorithms
	3 Proposed Prediction Models
	4 Experimental Results
		4.1 Data Usage
		4.2 Solar Radiation Prediction
	5 Conclusions
	References
A Recommender System for Insurance Packages Based on Item-Attribute-Value Prediction
	1 Introduction
	2 Preparation Steps
		2.1 Background Knowledge
		2.2 Data Preparation
		2.3 A Review of Techniques
	3 INSUREX Approach
		3.1 Overall Flow
		3.2 Item-Attribute-Value Prediction
		3.3 Recommendation Model
		3.4 Evaluation Method
	4 Result and Evaluation
		4.1 Analytical Results
		4.2 Proof-of-Concept (POC) Application
		4.3 Discussion
	5 Conclusion
	References
Application of Recommendation Systems Based on Deep Learning
	1 Introduction
	2 Deep Learning
		2.1 Autoencoder
		2.2 Restricted Boltzmann Machine
		2.3 Deep Belief Network
		2.4 Convolutional Neural Network
		2.5 Recurrent Neural Network
	3 Application of Deep Learning in Content-Based Recommendation System
		3.1 Method Based on Multilayer Perceptron
		3.2 Method Based on Convolutional Neural Network
		3.3 Methods Based on Other Deep Learning Models
	4 Application of Deep Learning in Collaborative Filtering Recommendation System
		4.1 Collaborative Filtering Based on Restricted Boltzmann Machine
		4.2 Collaborative Filtering Based on Autoencoder
		4.3 Collaborative Filtering Based on Convolutional Neural Network
		4.4 Collaborative Filtering Based on Recurrent Neural Network
		4.5 Collaborative Filtering Based on Other Deep Learning Models
	5 Application of Deep Learning in Hybrid Recommendation System
		5.1 Hybrid Recommendation Method Based on Autoencoder
		5.2 Hybrid Recommendation Method Based on Other Deep Learning Models
	6 Conclusion
	References
Data Reduction with Distance Correlation
	1 Introduction
	2 Related Work
	3 Methodology
	4 Results and Analysis
		4.1 Analysis
	5 Conclusion
	References
Intelligent and Contextual Systems
A Subtype Classification of Hematopoietic Cancer Using Machine Learning Approach
	1 Introduction
	2 Methodology
		2.1 Dataset
		2.2 Principal Component Analysis for Dimension Reduction
		2.3 Synthetic Minority Oversampling Technique
		2.4 Classification Algorithms
	3 Experiment and Result
	4 Conclusion
	References
Design of Text and Voice Machine Translation Tool for Presentations
	1 Introduction
	2 Background and Related Work
		2.1 Sequence to Sequence Model
		2.2 Speech Recognition
		2.3 Machine Translate
		2.4 Text to Speech
	3 System Description
		3.1 Speech Recognition
		3.2 Machine Translation
		3.3 Text to Speech
	4 Experiments and Results
		4.1 Vietnamese-English Translation
		4.2 English-Vietnamese Translation
	5 Conclusion
	References
Hybrid Approach for the Semantic Analysis of Texts in the Kazakh Language
	1 Introduction
	2 Related Works
	3 A Semantic Analysis Based an Algorithm for Extracting Annotation and Keywords
	4 Application of Approaches and Experimental Results
	5 Conclusion and Future Work
	References
Natural Language Processing
Short Text Clustering Using Generalized Dirichlet Multinomial Mixture Model
	1 Introduction
	2 Background
		2.1 Latent Dirichlet Allocation
		2.2 Gibbs Sampling for Dirichlet Multinomial Model
	3 Proposed Model
		3.1 Generalized Dirichlet Multinomial Mixture Model
		3.2 Collapsed Gibbs Sampling for Generalized Dirichlet Multinomial Mixture Model
	4 Experimental Results
		4.1 Short-Text Datasets and Preprocessing
		4.2 Evaluation Metrics Used
		4.3 Comparison of Gibbs Sampling Algorithms
		4.4 Influence of K
		4.5 Influence of the Number of Iterations
		4.6 Performance Given the Parameter
	5 Conclusion
	References
Comparative Study of Machine Learning Algorithms for Performant Text Analysis in a Real World System
	1 Introduction
		1.1 Motivation
	2 Problem Statement
	3 Related Work
	4 Data Processing
		4.1 Data Collection
		4.2 Data Preprocessing
		4.3 Word Cloud
		4.4 Keyword Extraction
		4.5 Data Mapping and Frequency
	5 Two-Class Neural Network
		5.1 Training the Two-Class Neural Network
	6 Two-Class Averaged Perceptron
		6.1 Training of Two-Class Averaged Perceptron
	7 Two-Class Boosted Decision Tree
		7.1 Training the Two-Class Boosted Decision Tree
	8 Experimental and Statistical Results
		8.1 Web Service Deployment and Prediction
	9 Conclusion and Future Work
	References
Contour: Penalty and Spotlight Mask for Abstractive Summarization
	1 Introduction
	2 Related Work
	3 Contour: Penalty and Spotlight Mask
		3.1 Penalty
		3.2 Spotlight
		3.3 Contour
	4 Experiments and Results
		4.1 Datasets
		4.2 Implementation Details
		4.3 Evaluate Metrics
		4.4 Comparison Basis
		4.5 Result
	5 Conclusion
	References
Heterogeneous Information Access System with a Natural Language Interface in the Context of Organization of Events
	1 Introduction
	2 Related Work
	3 Architecture
	4 Evaluation
		4.1 Functional Test
		4.2 Performance Analysis
		4.3 Tests Performed and Results Obtained
	5 Conclusions
	References
Network Systems and Applications
Single Failure Recovery in Distributed Social Network
	1 Introduction
	2 Related Work
	3 Problem Statement
	4 Proposed Methodology
		4.1 Framework
		4.2 Approach for Simulation
	5 Experimental Study
		5.1 System Specification
		5.2 Result
	6 Conclusion
	References
Improving Reading Accuracy in Personal Information Distribution Systems Using Smartwatches and Wireless Headphones
	1 Introduction
	2 Research Backgrounds
		2.1 Spread of the Internet
		2.2 Personalized Information Distribution System PINOT
		2.3 Smartphone Penetration
		2.4 Smartphone Version PINOT
		2.5 Problem of Smartphone Version PINOT
		2.6 Widespread Use of Smartwatches
		2.7 Smartwatch Cooperation PINOT
		2.8 Problem of Smartwatch Cooperation PINOT
	3 Proposed System
	4 Evaluation and Improvement of the Proposed System
		4.1 Experimental Results and Discussion
		4.2 Improvement of Misreadings
	References
Cyber Security Aspects of Digital Services Using IoT Appliances
	1 Introduction
	2 Security Threats to Communication Networks
		2.1 Botnets and Ransomware
		2.2 Threats to Applications of IoT
	3 Security Aspects of Customer Premises Network
		3.1 Communication Protocols’ Aspects
		3.2 Modeling of Threats and Incidents
	4 Security Testing and Validation
		4.1 Testing Environment
		4.2 Blockchain Database in a Sample IoT Application
	5 Concluding Remarks
	References
Computational Imaging and Vision
Simple Methodology for Eye Gaze Direction Estimation
	1 Introduction
	2 Detection of Facial Features
	3 Methodology
		3.1 Eye Gaze Direction Model
		3.2 Features for Gaze Direction Estimates
		3.3 Neural Network for Gaze Direction
	4 Feasibility-Study Results
		4.1 Dataset Building
		4.2 Neural Network Training
	5 Summary
	References
Dropout Regularization for Automatic Segmented Dental Images
	1 Introduction
	2 Related Work
	3 Methods and Techniques
		3.1 Dataset Preparation - Augmentation
		3.2 Pre-processing
		3.3 Model Description
		3.4 Dropout on a Neural Network
		3.5 Rectified Linear Unit Activation Function-RELU
	4 Experimental Results and Discussion
	5 Conclusion
	References
Towards Exploiting Convolutional Features for Remote Sensing Images Scene Classification
	1 Introduction
	2 Related Work
	3 Foundational Concepts
		3.1 Neurons and Neural Networks
		3.2 ConvNet Components
	4 Strategies for Exploiting ConvNets
		4.1 Full-Trained Networks
		4.2 Fine-Tuned Network
		4.3 Convolutional Neural Neworks as Feature Extractors
	5 The Proposed Method
	6 Experimentation
		6.1 Datasets
		6.2 Experiment Setting
		6.3 Evaluation Metrics for High-Level Feature Representation on Effectiveness and Robustness
	7 Results and Discussion
	8 Conclusions
	References
Contrast Enhancement in Deep Convolutional Neural Networks for Segmentation of Retinal Blood Vessels
	1 Introduction
		1.1 Related Work
	2 Method
		2.1 Data Preparation
		2.2 Pre-processing
		2.3 Architecture
	3 Experiments
		3.1 Experimental Setup and Training
		3.2 Evaluation Metrics
	4 Results and Discussion
	5 Conclusion
	References
Real-Time Social Distancing Alert System Using Pose Estimation on Smart Edge Devices
	1 Introduction
	2 Literature Review
		2.1 Deep Learning for Human Pose Estimation
		2.2 Social Distance Alert System Using Computer Vision
	3 Methodologies
		3.1 Key Frame Pre-processing
		3.2 Pose Estimation for Edge Devices Using PoseNet Model
		3.3 Object Distance Estimation
	4 Experiment
	5 Conclusion and Future Work
	References
Automatic Video Editor for Reportages Assisted by Unsupervised Machine Learning
	1 Introduction
	2 Background: Existent Voice Transcription Technologies
		2.1 Trends in Natural Language Processing
	3 Methodology
		3.1 Processing the Input Files and Preparing the File with Results
		3.2 Processing the Input Files and Preparing the File with Results
		3.3 Transcribing the Audio Clips and Discovering Topics
		3.4 Fetching and Preparing the Sentences and Creating an LDA Model
		3.5 Exporting Video and XML
		3.6 User Interface
	4 Results
		4.1 Comparing the Results in Relation to the Complexity of the Input Material
		4.2 Potential for Further Improvement
	5 Conclusion
	References
Residual Attention Network vs Real Attention on Aesthetic Assessment
	1 Introduction
	2 Methodology
	3 Results and Discussions
	4 Conclusion
	References
ICDWiz: Visualizing ICD-11 Using 3D Force-Directed Graph
	1 Introduction
	2 Backgrounds and Related Works
		2.1 Medical Classification
		2.2 Unified Medical Language System (UMLS) ch27Bodenreider2004
		2.3 Visualization Technique and Related Works
	3 Proposed Approach
		3.1 Data Preparation for the ICDWiz Database
		3.2 Visualization Construction
	4 Experimental Result
	5 Conclusion
	References
Decision Support and Control Systems
An Integrated Platform for Vehicle-Related Services and Records Management Using Blockchain Technology
	1 Introduction
	2 Proposed Approach
		2.1 Stakeholders
		2.2 Citizen Services
		2.3 Regional Transport Offices (RTOs) Activities
		2.4 Roles of Insurance Companies
		2.5 Pollution Control Testing Centers (PCTCs) Activities
		2.6 Activities of Traffic Police
		2.7 Access Control
	3 Proof of Concept
	4 Experimental Results
	5 Related Work and Comparative Analysis
	6 Conclusion
	References
Increase Driver’s Comfort Using Multi-population Evolutionary Algorithm
	1 Introduction
	2 Model of Investigations
	3 Optimization Problem
	4 Optimization Algorithm
	5 Calculation Results
	6 Conclusions
	References
The Impact of Cybersecurity on the Rescue System of Regional Governments in SmartCities
	1 Introduction
	2 Strategies and Regional Authorities in Smart Cities
	3 Security Components and Smart Technologies in Smart Cities
	4 Positive and Negative Impacts of the Use of Smart Technologies in Security Components in the Coordination of Regional Authorities in Smart Cities
		4.1 Positive Effects
		4.2 Negative Effects
	5 Conclusion
	References
A General Modeling Frameworkpg for Robust Optimization
	1 Introduction
	2 Basic Concepts
	3 Solution Methods
		3.1 Robust Linear Programming
		3.2 Min-Max Regret Linear Programming
		3.3 Min-Max Regret Integer Programming
	4 Examples
	5 Conclusions
	References
Unsupervised Barter Model Based on Natural Human Interaction
	1 Introduction
		1.1 General Agent Model
		1.2 The Walrasian Equilibrium as a Classical Barter Model
		1.3 Recent Works
	2 The Barter Mathematical Function Between Two Agents
	3 Results and Discussion
		3.1 Unsupervised Barter Scenarios Based on Natural Human Exchange of Goods
	4 Conclusions
	References
Data Modelling and Processing for Industry 4.0
Predictive Maintenance for Sensor Enhancement in Industry 4.0
	1 Introduction
	2 Materials and Methods
		2.1 Algorithm and Pipeline
		2.2 ARIMA Model
		2.3 Holt-Winters Model
		2.4 Neural Networks
		2.5 Distributed Data Science and Machine Learning
	3 Background
		3.1 Sliding Window and Distance-Based Methods
		3.2 Forecasting Methods
	4 Experimental Results
		4.1 Distance-Based Methods – Dynamic Time Warping
		4.2 Sliding Window and Distance-Based Methods – Matrix Profile
		4.3 Forecasting Methods
	5 Conclusion
	References
Cloud Based Business Process Modeling Environment – The Systematic Literature Review
	1 Introduction
	2 Business Process Lifecycle and Business Process Management
	3 Business Processes Modeling – Classic Approach
		3.1 BPMN
		3.2 Other Solutions
	4 Cloud Environment for Business Processes Modeling
		4.1 Business Process as a Service
		4.2 Elastic Business Process in BPMaaS
	5 Research Methods and Findings
		5.1 The Research Hypotheses
		5.2 The Research Methodology - A Systematic Literature Review
		5.3 Research Findings
	6 Conclusions
	References
Terrain Classification Using Neural Network Based on Inertial Sensors for Wheeled Robot
	1 Introduction
	2 Experiment
		2.1 Wheeled Robot
		2.2 Modules of the Terrain
		2.3 The Course of the Experiment
		2.4 Data Preparation
		2.5 Vector of Features
	3 The Neural Network
		3.1 Network Parameters
		3.2 Network Performance
	4 Conclusions
	References
Author Index