Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing (Studies in Computational Intelligence, 1012)

Preface
Organization
	International Program Committee
	General Chair
	Conference Chairs
	Program Chairs
	Registration Co-chair
	Local Arrangements Chair
	Publicity Co-chair
	Finance Chair
	Publication Chairs
	Committee Members
	SNPD 2021-Fall Program Committee
Contents
Contributors
Segmentation of Spinal MRI Images and New Compression Fracture Detection
	Abstract
	1 Introduction
	2 Methodology
		2.1 Segmentation of Spinal
			2.1.1 Data Augmentation
			2.1.2 U-Net Architecture
			2.1.3 Binary Cross Entropy
			2.1.4 Dice Loss
		2.2 Vertebra Regions Segmentation
			2.2.1 False Vertebra Regions Removing
			2.2.2 Noise Removing Operation
			2.2.3 Hole Filling Operation
			2.2.4 Opening Operation
		2.3 Gamma Correction
		2.4 Symptoms Classification
			2.4.1 ResNet50 Architecture
			2.4.2 Oversampling + Undersampling
			2.4.3 Class Weight
	3 Experiment Result
		3.1 Segmentation Result
		3.2 Classification Result
			3.2.1 Model Comparing
			3.2.2 Data Imbalance Problem
			3.2.3 Classification Methods
			3.2.4 Gamma Correction
			3.2.5 Results Presentation
		3.3 Result Analysis
			3.3.1 Normal Vertebra is Misjudged as Old Compression Fracture
			3.3.2 Distinction Between New and Old Compression Fractures
			3.3.3 MRI Image of No Screw Features
	4 Conclusions
	References
Key Issues for Digital Factory Designing and Planning: A Survey
	1 Introduction
	2 Research on Key Issues of DF from Perspectives of Networking, Precision, Automation and Digitalization
	3 Research on Key Technologies of 5G Network, Time Sensitive Network and Digital Twin
	4 Conclusion
	References
Concentration-Based Robot Control Method with FPGA
	1 Introduction
	2 Method Implementation
		2.1 Framework of the Control Method
		2.2 Hardware of the Control Method
		2.3 Software of the Control Method
		2.4 Android App
	3 Method Performance Test
	4 Conclusion
	References
SolarWinds Software Supply Chain Security: Better Protection with Enforced Policies and Technologies
	Abstract
	1 Introduction
	2 SolarWinds Software Supply Chain Breaches
	3 SolarWinds Software Supply Chain Attack on Policies
		3.1 Market Incentives for Security
		3.2 ‘Additive’ Security to ‘Reductive’ Security
		3.3 No Updates if Unnecessary
		3.4 Tool Development for Customers to Evaluate Security on Updates
		3.5 Ubiquitous Use of Strong Encryption and Regular Movement of Sensitive Data on Clouds and Fogs
		3.6 Attempts to Apply AI and Quantum-Based Approach
	4 Ways to Detect and Protect Against
		4.1 Software Composition Analysis (SCA)
		4.2 CHIRP (CISA Hunt and Incident Response Program)
		4.3 The SootDiff System
		4.4 The In-toto System
	5 Conclusion and Future Work
	References
A Data Hiding Technology by Applying Interpolation in Extended Local Binary Pattern
	Abstract
	1 Introduction
	2 Related Method
	3 Proposed Method
	4 Experimental Results
	5 Conclusions
	Acknowledgements
	References
Insect Species Identification System Based on Deep Learning
	Abstract
	1 Introduction
		1.1 Background and Motivation
		1.2 Purpose
		1.3 Thesis Structure
	2 Research Methods and Procedures
		2.1 RGB Image and YUV Image Conversion
			2.1.1 Gamma Correction
		2.2 Object Detection Model
			2.2.1 Basic Concepts of YOLO
			2.2.2 Input and Output of Detection Model
		2.3 Contrast Limited Adaptive Histogram Equalization (Clahe)
		2.4 Species Identification Model
			2.4.1 GoogLeNet Inception-v4 Structure
			2.4.2 Input and Output of the Species Identification Model
		2.5 RGB Normalization
		2.6 Combined Model
	3 Experimental Results and Discussion
		3.1 Experimental Image Dataset
		3.2 Experimental Results of Object Detection
			3.2.1 Evaluation Criteria for Experimental Results
			3.2.2 Experimental Results of the Object Detection Model
			3.2.3 Discussion of Object Detection Results
		3.3 Experimental Results of Species Identification
			3.3.1 Evaluation Criteria for Experimental Results
			3.3.2 Experimental Results of the Primary Identification Model
			3.3.3 Experimental Results of the Secondary Identification Model
			3.3.4 Experimental Results of Combined Model of Primary and Secondary Identification Models
			3.3.5 Discussion of Species Identification Results
	4 Conclusions and Future Prospects
	References
Study of DIFA Based Learning Data Generating Methodology for Malware Detection
	1 Introduction
	2 IT Trends and Analysis Environment
		2.1 Information Service Trends and Security Environment
		2.2 Malware Trends
		2.3 Malware Analysis Environment
	3 Machine Learning in Malware Analysis
		3.1 Advantage of Applying Machine Learning to Analysis Malware
		3.2 Tries of Applying Machine Learning About Analysis Malware
	4 DIFA Learning Data Transforming Methodology for Application Misuse Detection
		4.1 Machine Learning Requirement for Analysis Malware
		4.2 Data Transforming Methodology for Malware Analysis Detection
		4.3 Data Generation Method at Kernel Event for Malware Analysis
	5 Implement
		5.1 Implement Environment
		5.2 Detection Rate by Learning Model
		5.3 Detection Rate Analysis
	6 Conclusion
	References
A Study on the Effect of Educational Service Quality on Career Decision-Making Self-efficacy
	Abstract
	1 Introduction
	2 Theoretical Background
		2.1 Educational Service Quality
		2.2 Learning Flow and Educational Satisfaction
		2.3 Academic Achievement and Career Decision-Making Self-efficacy
	3 Research Model and Hypotheses
		3.1 Research Model and Hypotheses
		3.2 Measurement and Analysis Method
	4 Empirical Analysis
		4.1 Data Collection and the Characteristics of Samples
		4.2 Analysis of the Measurement Model
		4.3 Structural Model and Hypothesis Verification
	5 Conclusion
		5.1 Study Results
		5.2 Implications of the Study
	References
Nanotechnology Performance Analysis Using Topic Modeling and Social Network Analysis: Focusing on NTIS Data in Korea (2015–2019)
	Abstract
	1 Introduction
	2 Theoretical Background and Hypotheses
		2.1 Nano Technology
		2.2 Nano Technology Trend Analysis
	3 Research Method
		3.1 Topic Modeling Analysis
		3.2 Text Network Analysis
	4 Empirical Analysis
		4.1 Topic Modeling Analysis
		4.2 Text Network Analysis
			4.2.1 Analysis of Key Factors Affecting Nanotechnology Performance
			4.2.2 Analysis of Major Nanotechnology Achievements by Period
			4.2.3 Analysis of Changes in Nanotechnology Performance Fields According to Time
	5 Conclusions
	Acknowledgment
	References
A Study on Korea TV Drama Ratings: Programming and Marketing Strategies
	Abstract
	1 Introduction
	2 Previous Studies
	3 Data
	4 Method
		4.1 Multi Linear Regression
	5 Experiment Design
		5.1 Variables
	6 Results
	7 Conclusion
	Acknowledgement
	References
Identifying the Public’s Changing Concerns During a Global Health Crisis: Text Mining and Comparative Analysis of Tweets During the COVID-19 Pandemic
	Abstract
	1 Introduction
	2 Methods
		2.1 Data Acquisition
		2.2 Text Mining
	3 Results
		3.1 #COVID19: February 28–March 5, 2020
		3.2 #Coronavirus: February 28–March 5, 2020
		3.3 #COVID19: June 15–June 21, 2020
		3.4 #Coronavirus: June 15–June 21, 2020
	4 Discussion
	5 Conclusion/Limitation and Future Research
	Acknowledgement
	References
The Discovery of Historical Transition in Aesthetic Notions Through Changes in Co-occurrence Words Mainly Used in Waka Poetry in Three Major Poetry Anthologies
	Abstract
	1 Introduction
	2 Related Works
		2.1 Study on the Framework of Kago
		2.2 Study on the History of Waka Expressions
		2.3 Study on Classical Japanese Poetry by Pattern Extraction
		2.4 Study on Waka Poetry Using Character N-Grams
		2.5 Position of Our Study
	3 The Discovery of Historical Transition in Aesthetic Notions Through Changes in Co-occurrence Words of Kago
		3.1 Overview
		3.2 Data Sets of Waka Poetry
		3.3 Data Set of Kago
		3.4 Morphological Analysis Function
		3.5 Kago Extraction Function
		3.6 Latent Meaning of Kago Extraction Function
	4 Experiment
		4.1 Experiment Environment
		4.2 Experiment 1 (The Result of the Proportion of Kago Used in Each of the Three Major Anthologies)
		4.3 Experiment 2 (The Results of the Verification of the Changes in Co-occurrence Words of Kago in Each of the Three Major Anthologies)
		4.4 Experiment 3 (The Results of Changes in the Subjects on Which Waka Poets Put Their Thoughts)
		4.5 Discussion
	5 Conclusion
	Acknowledgments
	References
Facial Expression Recognition Using Deep Learning Methods
	Abstract
	1 Introduction
	2 Related Work
	3 Methodology
		3.1 Preprocessing
		3.2 Data Augmentation
		3.3 Network Ensembles
		3.4 Implementation
	4 Experimental Study
		4.1 Data Used
		4.2 Results
	5 Conclusions
	Acknowledgement
	References
Author Index