Trends and applications in knowledge discovery and data mining : PAKDD 2020 workshops, DSFN, GII, BDM, LDRC and LBD, Singapore, May 11-14, 2020 : Revised selected papers

This book constitutes the thoroughly refereed post-workshop proceedings of the workshops that were held in conjunction with the 24th Pacific-Asia Conference on Knowledge Discovery and Data Mining, PAKDD 2020, in Singapore, Singapore, in May 2020. The 17 revised full papers presented were carefully reviewed and selected from a total of 50 submissions. The five workshops were as follows: · First International Workshop on Literature-Based Discovery (LBD 2020) · Workshop on Data Science for Fake News (DSFN 2020) · Learning Data Representation for Clustering (LDRC 2020) · Ninth Workshop on Biologically Inspired Techniques for Data Mining (BDM · 2020) · First Pacific Asia Workshop on Game Intelligence & Informatics (GII 2020) Preface 6 Organization 7 Contents 9 First International Workshop on Literature-Based Discovery (LBD 2020) 11 Matching Biomedical Ontologies with Compact Evolutionary Algorithm 12 1 Introduction 13 2 Preliminaries 13 2.1 Biomedical Ontology Matching Problem 13 2.2 Similarity Measure on Biomedical Concept 14 3 Compact Evolutionary Algorithm 14 4 Experimental Results 15 4.1 Comparison on Alignment's Quality and Runtime 17 5 Conclusion 18 References 19 Towards Expert Preference on Academic Article Recommendation Using Bibliometric Networks 20 1 Introduction 20 2 Materials and Methods 21 2.1 Domain Expert Recommendation Analysis 21 2.2 Computational Model 23 3 Results 25 4 Conclusion 27 References 27 Research in Progress: Explaining the Moderating Effect of Massively Multiplayer Online (MMO) Games on the Relationship Between Flow and Game Addiction Using Literature-Based Discovery (LBD) 29 1 Introduction 29 2 The Moderating Effect of MMO 30 3 Research Methodology 31 4 Data Analysis 32 5 Discussion 34 6 Conclusion and Future Research 34 Appendix 35 References 37 A Network Approach for Mapping and Classifying Shared Terminologies Between Disparate Literatures in the Social Sciences 39 1 Introduction 39 2 Data and Methods 40 3 Results 42 4 Discussion and Conclusion 46 References 48 Towards Creating a New Triple Store for Literature-Based Discovery 50 1 Introduction 50 2 Our Contribution: An Extended Knowledge Base 52 2.1 New Types of Predicates, Entities and Relations 52 2.2 Data Sources 53 2.3 Database Design 53 3 Methods 54 3.1 Overview of the Approach 54 3.2 Creation of a Corpus for Distant Supervision 55 3.3 Distantly Supervised Relation Extraction 55 4 Conclusion and Difficulties 55 References 56 Who Is Who in Literature-Based Discovery: Preliminary Analysis 60 1 Introduction 60 2 Methods 61 2.1 Collection of Bibliographic Data 61 2.2 Data Analysis 61 3 Results 62 3.1 Publication Evolution over the Years 62 3.2 Most Productive Authors 63 3.3 Most Productive Countries 63 3.4 Most Relevant Journals 63 3.5 Publication Hallmarks 66 4 Discussion 66 References 67 Workshop on Data Science for Fake News (DSFN 2020) 69 Adversarial Deep Factorization for Recommender Systems 70 1 Introduction 70 2 Adversarial Deep Factorization 72 3 Experiments 75 3.1 Datasets and Evaluation Metric 75 3.2 Summary of Experimental Results 76 4 Conclusion 77 References 77 Detection of Spammers Using Modified Diffusion Convolution Neural Network 79 1 Introduction 79 2 Problem Definition and Model Introduction 80 2.1 Problem Definition 80 2.2 DCNN Model 81 2.3 DCNN+CR Model 82 3 Experiment 83 3.1 Data Processing 83 3.2 Comparative Baselines and Experimental Results 84 4 Conclusion 85 References 85 Dynamics of Online Toxicity in the Asia-Pacific Region 87 1 Introduction 87 2 Literature Review 88 2.1 Online Toxicity 88 2.2 Applying Topic Modeling to Social Media Data 89 3 Methodology 89 4 Results and Discussion 90 4.1 Toxicity Analysis 90 4.2 Topic Modeling 91 5 Conclusion and Future Work 94 References 94 Workshop on Learning Data Representation for Clustering (LDRC 2020) 95 Predictive K-means with Local Models 96 1 Introduction 96 2 Turning the K-means Algorithm Predictive 97 3 Comparison with Supervised Algorithm 102 3.1 The Chosen Set of Classifiers 102 3.2 Experimental Protocol 104 3.3 Results 104 4 Conclusion and Perspectives 107 References 107 Automatic Detection of Sexist Statements Commonly Used at the Workplace 109 1 Introduction 109 2 Related Works 111 3 Dataset Description 112 4 Description of the Classification Method 114 5 Experimental Results 117 6 Conclusion and Future Works 118 References 119 Hybrid Dimensionality Reduction Technique for Hyperspectral Images Using Random Projection and Manifold Learning 121 1 Introduction 121 2 Hybrid Band Extraction Methodology 123 2.1 Random Projection 124 2.2 Kernel PCA 124 3 Experiments 125 3.1 Performance Evaluation Measures 126 3.2 Experiment Setup 127 4 Result Analysis and Discussion 128 5 Conclusion 130 References 131 Clustering Spatial Data via Mixture Models with Dynamic Weights 133 1 Introduction 133 2 Hidden Logistic Random Field Mixture Model 134 2.1 Spatial Data Segmentation 137 2.2 Parameter Estimation Algorithm 137 3 Experiments 139 4 Conclusion 142 References 142 Ninth Workshop on Biologically Inspired Techniques for Data Mining (BDM 2020) 144 Temporal Convolutional Network Based Transfer Learning for Structural Health Monitoring of Composites 145 1 Introduction 145 2 Methodology 147 2.1 Causal Dilated Convolution 147 2.2 Residual Temporal Block 148 2.3 TCN Based Transfer Learning (TCN-trans): 149 3 Results 149 3.1 Data 149 3.2 Preprocessing and Training 150 3.3 TCN Parameter Selection 151 3.4 Effect of Input Parameters on Delamination Prediction 151 3.5 TCN-trans: TCN Based Transfer Learning for Delamination Prediction 153 4 Conclusions 154 References 155 Dilated Convolutional Recurrent Deep Network with Transfer Learning for Remaining Useful Life Prediction 157 1 Introduction 157 2 Methodology 159 2.1 Dilated Component 159 2.2 Recurrent Component 161 2.3 DCNN-LSTM with Transfer Learning 161 3 Experiment Results 162 3.1 DCNN-LSTM Performance Evaluation 162 3.2 DCNN-LSTM with Transfer Learning 164 4 Conclusions 167 References 168 Generative Modeling for Synthesis of Cellular Imaging Data for Low-Cost Drug Repurposing Application 169 1 Introduction 169 2 Methods 171 2.1 Overview of the Framework 171 2.2 HCS Imaging Data 172 2.3 Generative Modeling: HCS Image Synthesis 172 2.4 Pre-trained CNN Based Feature Extraction 174 2.5 GMM Based Phenotype Clustering 175 3 Results 175 3.1 Generation of HCS Images: Comparison of Generative Models 176 3.2 Clustering Performance: Evaluation of Synthetic Images 176 3.3 Grouping of Similar Drugs: Improved Confidence 179 4 Discussion 179 References 180 First Pacific Asia Workshop on Game Intelligence and Informatics (GII 2020) 182 A Study of Game Payment Data Mining: Predicting High-Value Users for MMORPGs 183 1 Introduction 183 2 Datasets 184 3 Data Analysis 185 3.1 Preliminary 185 3.2 Analysis Payment Behavior Underlying CLVs 185 3.3 Analysis of Potential High-Value Users 188 3.4 Summary 188 4 Proposed Model 188 4.1 Notations 188 4.2 Models 189 4.3 Model Inference 189 5 Experiments and Results 191 5.1 Experimental Settings 191 5.2 Performance 191 6 Related Works 193 7 Conclusions 194 References 194 Author Index 195 Front Matter ....Pages i-x Front Matter ....Pages 1-1 Matching Biomedical Ontologies with Compact Evolutionary Algorithm (Xingsi Xue, Pei-Wei Tsai)....Pages 3-10 Towards Expert Preference on Academic Article Recommendation Using Bibliometric Networks (Yu Zhang, Min Wang, Morteza Saberi, Elizabeth Chang)....Pages 11-19 Research in Progress: Explaining the Moderating Effect of Massively Multiplayer Online (MMO) Games on the Relationship Between Flow and Game Addiction Using Literature-Based Discovery (LBD) (Wendy Hui, Wai Kwong Lau)....Pages 20-29 A Network Approach for Mapping and Classifying Shared Terminologies Between Disparate Literatures in the Social Sciences (Cristian Mejia, Yuya Kajikawa)....Pages 30-40 Towards Creating a New Triple Store for Literature-Based Discovery (Anna Koroleva, Maria Anisimova, Manuel Gil)....Pages 41-50 Who Is Who in Literature-Based Discovery: Preliminary Analysis (Andrej Kastrin, Dimitar Hristovski)....Pages 51-59 Front Matter ....Pages 61-61 Adversarial Deep Factorization for Recommender Systems (Ziqing Chen, Yiwei Zhang, Zongyang Li)....Pages 63-71 Detection of Spammers Using Modified Diffusion Convolution Neural Network (Hui Li, Wenxin Liang, Zihan Liao)....Pages 72-79 Dynamics of Online Toxicity in the Asia-Pacific Region (Thomas Marcoux, Adewale Obadimu, Nitin Agarwal)....Pages 80-87 Front Matter ....Pages 89-89 Predictive K-means with Local Models (Vincent Lemaire, Oumaima Alaoui Ismaili, Antoine Cornuéjols, Dominique Gay)....Pages 91-103 Automatic Detection of Sexist Statements Commonly Used at the Workplace (Dylan Grosz, Patricia Conde-Cespedes)....Pages 104-115 Hybrid Dimensionality Reduction Technique for Hyperspectral Images Using Random Projection and Manifold Learning (Alkha Mohan, M. Venkatesan)....Pages 116-127 Clustering Spatial Data via Mixture Models with Dynamic Weights (Allou Samé)....Pages 128-138 Front Matter ....Pages 139-139 Temporal Convolutional Network Based Transfer Learning for Structural Health Monitoring of Composites (Shaista Hussain, Luu Trung Pham Duong, Nagarajan Raghavan, Mark Hyunpong Jhon)....Pages 141-152 Dilated Convolutional Recurrent Deep Network with Transfer Learning for Remaining Useful Life Prediction (Jing Yang Lee, Ankit K. Das, Shaista Hussain, Yang Feng)....Pages 153-164 Generative Modeling for Synthesis of Cellular Imaging Data for Low-Cost Drug Repurposing Application (Shaista Hussain, Ayesha Anees, Ankit K. Das, Binh P. Nguyen, Mardiana Marzuki, Shuping Lin et al.)....Pages 165-177 Front Matter ....Pages 179-179 A Study of Game Payment Data Mining: Predicting High-Value Users for MMORPGs (Junxiang Jiang)....Pages 181-192 Back Matter ....Pages 193-193