Knowledge Management and Acquisition for Intelligent Systems: 17th Pacific Rim Knowledge Acquisition Workshop, PKAW 2020, Yokohama, Japan, January ... (Lecture Notes in Computer Science, 12280)

This book constitutes the proceedings of the 17th International Workshop on Knowledge Management and Acquisition for Intelligent Systems, PKAW 2020, held in Yokohama, Japan, in January 2021. The 10 full papers and 5 short papers included in this volume were carefully reviewed and selected from 28 initial submissions. PKAW primarily focuses on the multidisciplinary approach of the human-driven and data-driven knowledge acquisition, which is the key concept that has remained unchanged since the workshop has been established. Preface Organization Contents Accelerating the Backpropagation Algorithm by Using NMF-Based Method on Deep Neural Networks 1 Introduction 2 Training of DNN 2.1 Backpropagation 2.2 NMF-Based Algorithm 3 Proposed Method 4 Experimental Results 4.1 Preliminary Experiments on the BP and the NMF-Based Algorithm 4.2 Performance Evaluation 5 Conclusion References Collaborative Data Analysis: Non-model Sharing-Type Machine Learning for Distributed Data 1 Introduction 2 Distributed Data 3 Collaborative Data Analysis 3.1 Basic Concept 3.2 Derivation of the Proposed Method 3.3 Practical Operation Strategy Regarding Privacy and Confidentiality Concerns 4 Numerical Experiments 4.1 Experiment I: For Artificial Data 4.2 Experiment II: vs. Number of Parties 4.3 Experiment III: For Real-World Data 4.4 Remarks on Numerical Results 5 Conclusions References ERA: Extracting Planning Macro-Operators from Adjacent and Non-adjacent Sequences 1 Introduction 2 Background Theory 3 ERA 3.1 Overview 3.2 ERA Algorithm 3.3 Mining Procedure 3.4 Complexity Analysis. 4 Evaluation 4.1 Experimental Setup 4.2 Results of the ERA Algorithm 5 Discussion 6 Conclusion References Deep Neural Network Incorporating CNN and MF for Item-Based Fashion Recommendation 1 Introduction 2 Related Works 2.1 Outfit Recommendation 2.2 Item-Based Recommendation 3 Methodology 3.1 Offline Training Phase 3.2 Item Recommendation Phase 4 Experiments 4.1 Datasets 4.2 Evaluation Method 4.3 Comparison Methods 4.4 Performance 4.5 Online Evaluation 5 Conclusion References C-LIME: A Consistency-Oriented LIME for Time-Series Health-Risk Predictions 1 Introduction 2 Related Works 2.1 Risk Prediction Based on Electronic Health Records (EHRs) 2.2 Explainable Artificial Intelligence (XAI) 2.3 Summary 3 Proposed Method 3.1 Health-Risk Prediction Algorithm 3.2 Interpretation of Health-Risk Prediction by LIME 3.3 C-LIME: Consistently Explainable LIME 4 Evaluation 4.1 Accuracy of Health-Risk Prediction Model 4.2 Interpretation of Health Risk Prediction Model Using C-LIME 5 Health-Risk Prediction and Lifestyle Recommendation Service 6 Conclusions References Discriminant Knowledge Extraction from Electrocardiograms for Automated Diagnosis of Myocardial Infarction 1 Introduction 2 Our Approach 2.1 Overview 2.2 Modelling Spectral and Longitudinal Characteristics 3 Experiments 3.1 Datasets 3.2 Setup 3.3 Results and Discussion 4 Conclusions References Stabilizing the Predictive Performance for Ear Emergence in Rice Crops Across Cropping Regions 1 Introduction 2 Related Studies 3 Data 3.1 Cropping Records 3.2 Micro Climate Data 3.3 Partitioned Regions 4 Proposal - Engineering Variables 4.1 Clustering Regional Time Series Patterns 4.2 Deriving the Time Series Statistics 5 Predicting Procedure 5.1 Datasets 5.2 Executing Predictions 5.3 Evaluating Predictive Performance 6 Evaluation 7 Discussion 8 Conclusion References Description Framework for Stakeholder-Centric Value Chain of Data to Understand Data Exchange Ecosystem 1 Introduction 2 Data Exchange Ecosystem and Relevant Studies 3 Stakeholder-Centric Value Chain of Data 4 Experimental Details 5 Results and Discussion 5.1 Structural Characteristics of the Data Exchange Ecosystem 5.2 Knowledge Extraction from SVC 5.3 Limitations and Future Work 6 Conclusion References Attributed Heterogeneous Network Embedding for Link Prediction 1 Introduction 2 Related Work 3 The Proposed Model 3.1 Preliminaries 3.2 AHNE 4 Experimental Evaluation 4.1 Datasets 4.2 Experimental Settings 4.3 Experimental Results 5 Conclusion References Automatic Generation and Classification of Malicious FQDN 1 Introduction 2 Related Work 3 Criticism of String-Based Approach 4 FakePopular and Its Implication 5 Conclusion References Analyzing Temporal Change in LoRa Communication Quality Using Massive Measurement Data 1 Introduction 2 Related Works 3 Bus Location Management System and Its Log Data 4 Proposed Method 5 Analysis 5.1 Effects of Rainfall on RSSI 5.2 Effects of Fallen Snow on RSSI 6 Conclusion References Challenge Closed-Book Science Exam: A Meta-Learning Based Question Answering System 1 Introduction 2 Related Work 2.1 Standardized Science Exams 2.2 Meta-Learning 3 Meta-Classifier System 3.1 Few-Shot Question Classification 3.2 Model Agnostic Meta-Learning Method 4 Reasoning System 5 Experiments 5.1 Few-Shot Question Classification 5.2 Model Visualization 5.3 Question Answering with Few-Shot QC Information 5.4 Case Study 6 Conclusion References Identification of B2B Brand Components and Their Performance's Relevance Using a Business Card Exchange Network 1 Introduction 2 Related Work 3 Description of Dataset 3.1 Eight Company Score 3.2 Definition of Corporate Brand Score 3.3 Definition of Corporate Performance 4 RQ1: Is B2B Company Brand Related to Corporate Performance? 4.1 Discussion 5 RQ2: What Are the Components of a B2B Company's Brand Impression? 5.1 Supervised Topic Models 5.2 Relationship Between Each Latent Topic and Corporate Performance 5.3 Discussion 6 Conclusion References Semi-automatic Construction of Sight Words Dictionary for Filipino Text Readability 1 Introduction 2 Related Works 3 Extracting Seed Words from Storybooks 4 Expanding the Dictionary 4.1 Word Embeddings 4.2 Dictionary Entries 5 Discussion 5.1 Readability Assessment 5.2 Oral Reading Fluency 6 Conclusion and Future Work References Automated Concern Exploration in Pandemic Situations - COVID-19 as a Use Case 1 Introduction 2 Related Works 3 Automated Public Concern Detection 3.1 Data Pre-processing and Information Extraction 3.2 Deep Learning Models 3.3 Concern Extraction and Clustering 3.4 Concern Knowledge Graph 4 Experiments 4.1 Dataset 4.2 Results Analysis 4.3 Result Visualisation 5 Conclusion and Future Work References Author Index