Big data innovations and applications : 5th International Conference, Innovate-Data 2019, Istanbul, Turkey, August 26-28, 2019 : proceedings

This volume constitutes the refereed proceedings of the 5th International Conference on Big Data Innovations and Applications, Innovate-Data 2019, held in Istanbul, Turkey, in August 2019. The 15 revised full papers and 1 short paper presented in this volume were carefully reviewed and selected from 48 submissions. The papers are organized in topical sections on advances in big data systems; machine learning and data analytics; big data innovation and applications; security and risk analysis. Preface 6 Organization 8 Contents 11 Advances in Big Data Systems 13 Enabling Joins over Cassandra NoSQL Databases 14 Abstract 14 1 Introduction 14 2 Background 16 3 Query Optimization and Execution 18 3.1 Index-Nested Loops 20 3.2 Sort-Merge Join 21 4 Implementation and Evaluation 23 5 Conclusions and Discussion 26 Acknowledgements 27 References 27 MapReduce Join Across Geo-Distributed Data Centers*-10pt 29 1 Introduction 29 2 Background and Related Work 30 2.1 Background 30 2.2 Distributed Joins in MapReduce Environments 32 2.3 MapReduce in Geo-Distributed Environments 33 3 Hierarchical Hadoop Framework 34 4 Implementing MapReduce Join in H2F 37 5 Conclusion 41 References 41 Mobile Device Identification via User Behavior Analysis*-10pt 43 1 Introduction 43 2 Related Work 45 3 Methodology 46 3.1 Data Collection 46 3.2 Sensor Types 47 3.3 Data Preprocessing 49 3.4 Feature Extraction 49 3.5 Classification and Performance Metrics 50 4 Performance Evaluation 51 4.1 Accelerometer 52 4.2 Gyroscope 52 4.3 Accelerometer and Gyroscope 53 5 Conclusion and Future Work 56 References 57 A SAT-Based Formal Approach for Verifying Business Process Configuration 58 1 Introduction 58 2 Motivating Example and Background 59 2.1 Motivating Example: Configurable Business Process 60 2.2 Propositional Satisfiability 61 3 SAT-Based Business Process Configuration 62 3.1 Our Approach in a Nutshell 62 3.2 Formalizing Business Process Configuration Using SAT 63 4 Evaluation: SAT Problems Induced by Business Process 67 5 Related Work 69 6 Conclusion and Further Work 71 References 71 Machine Learning and Data Analytics 74 TSWNN+: Check-in Prediction Based on Deep Learning and Factorization Machine 75 Abstract 75 1 Introduction 75 2 Related Work 76 3 Feature Analysis and Feature Description 77 3.1 Feature Analysis 77 3.2 Feature Description 78 4 Proposed Model 79 4.1 Model Description 79 4.2 Loss Function 82 4.3 Implicit Data Expansion 82 5 Experiments and Results 84 5.1 Dataset 84 5.2 Evaluation Procedure and Performance Metric 85 5.3 Baselines 85 5.4 Results and Analysis 85 5.5 Performance Improvement 87 6 Conclusion 88 Acknowledgments 88 References 88 Deep Learning Based Sentiment Analysis on Product Reviews on Twitter 90 Abstract 90 1 Introduction 90 2 Related Work 92 3 Methodology 93 3.1 Dataset Collection and Preprocessing 93 3.2 Word Embeddings Based Feature Representation 94 3.3 Convolutional Neural Network Architecture 95 4 Experimental Procedure and Results 95 4.1 Experimental Procedure 95 4.2 Evaluation Measures 95 4.3 Experimental Results 96 5 Conclusion 99 References 99 A Cluster-Based Machine Learning Model for Large Healthcare Data Analysis*-10pt 102 1 Introduction 102 1.1 Technical Significance 103 1.2 Clinical Relevance 104 2 Data 104 2.1 Data Collection 104 3 Desired Output, EPIC-26 104 4 Method 105 4.1 Patient Clustering 107 4.2 Patient Classification 108 4.3 Extracting Important Features 109 4.4 Prediction of the EPIC-26 Actual Values 109 5 Results 109 5.1 Baseline 109 5.2 Evaluation Approach/Study Design 109 5.3 Case Study Results 112 6 Conclusions and Related Works 115 References 115 Satire Detection in Turkish News Articles: A Machine Learning Approach 117 Abstract 117 1 Introduction 117 2 Related Work 118 3 Methodology 119 3.1 Dataset Collection and Preprocessing 120 3.2 Feature Extraction Schemes 120 3.3 Classification Algorithms 121 4 Experimental Procedure and Results 122 4.1 Evaluation Measures 122 4.2 Experimental Procedure 122 4.3 Experimental Results 123 5 Conclusion 126 References 126 Big Data Innovation and Applications 128 Committee of the SGTM Neural-Like Structures with Extended Inputs for Predictive Analytics in Insurance 129 Abstract 129 1 Introduction 129 2 Review and Analysis of Existing Methods 130 3 Proposed Committee of the SGTM Neural-Like Structures with Extended Inputs 131 4 Modeling and Results 133 5 Comparison and Discussion 137 6 Conclusion 139 References 139 Game Analytics on Free to Play 141 Abstract 141 1 Introduction 141 2 Aspects of Game Analytics 142 2.1 Motivation to Use Game Analytic Tools for SME 143 2.2 Churn Prediction 143 2.3 Customer Lifetime Value Prediction 145 3 Conclusion 147 References 147 A Decentralized File Sharing Framework for Sensitive Data*-10pt 150 1 Introduction 150 1.1 Blockchain 151 1.2 IPFS 151 2 Related Work 152 3 A File Sharing Framework for Sensitive Data 153 3.1 Meta-data File 155 4 Evaluation 155 5 Future Work 156 6 Conclusion 156 References 156 The Information System for the Research in Carotid Atherosclerosis 158 1 Introduction 158 2 Atherosclerotic Plaque in B-image 159 3 Echogenicity Grade of Atherosclerotic Plaques 159 4 How Echo-Index Is Computed in Selected ROI 160 4.1 Core of the Algorithm 160 4.2 Two Approaches How to Select ROI 161 4.3 Echo-Index for Free-Hand ROI 164 5 Reproducibility of the Echo-Index Value in a Pilot Study 165 5.1 Achieved Results to the Reproducibility Assessment and Using Echo-Index 166 5.2 Basic Statistical Analysis of Computed Echo-Index Values 167 5.3 Role of the Echo-Index in Future Clinical Studies 168 6 Conclusions and Future Work 168 References 169 Security and Risk Analysis 170 Big Data Analytics for Financial Crime Typologies 171 Abstract 171 1 Introduction 171 2 Related Work 173 3 Generated Data Boundary Checks 174 4 Discussion 182 5 Conclusion 184 Acknowledgement 184 References 184 Development of a Model for Identifying High-Risk Operations for AML/CFT Purposes 185 Abstract 185 1 Introduction 185 2 Primary Data Processing 186 3 Data Grooming 189 4 Data Clustering 196 5 Conclusion 197 References 198 Monitoring System for the Housing and Utility Services Based on the Digital Technologies IIoT, Big Data, Data Mining, Edge and Cloud Computing 199 Abstract 199 1 Introduction 199 2 Problem Statement 200 2.1 Housing and Utility Services 200 2.2 The Industrial Internet of Things 200 2.3 Big Data 202 2.4 Edge, Fog and Cloud Computing 202 2.5 Data Mining 204 3 The Proposed Approach 205 3.1 Objects of the Housing and Utilities Complex 207 3.2 Object Cloud 208 3.3 Users 210 4 Conclusion 210 Acknowledgement 210 References 210 K-Means Method as a Tool of Big Data Analysis in Risk-Oriented Audit 212 Abstract 212 1 Introduction 212 2 K-Means Method as a Tool for Building an Audit Sample 212 3 Example of Application of the K-Means Method and Analysis of the Results in Comparison with the Continuous Test 217 3.1 Preparatory Stage 217 3.2 Application of the K-Means Method 219 References 222 Author Index 223 Front Matter ....Pages i-xii Front Matter ....Pages 1-1 Enabling Joins over Cassandra NoSQL Databases (Haridimos Kondylakis, Antonis Fountouris, Apostolos Planas, Georgia Troullinou, Dimitris Plexousakis)....Pages 3-17 MapReduce Join Across Geo-Distributed Data Centers (Giuseppe Di Modica, Orazio Tomarchio)....Pages 18-31 Mobile Device Identification via User Behavior Analysis (Kadriye Dogan, Ozlem Durmaz Incel)....Pages 32-46 A SAT-Based Formal Approach for Verifying Business Process Configuration (Abderrahim Ait Wakrime, Souha Boubaker, Slim Kallel, Walid Gaaloul)....Pages 47-62 Front Matter ....Pages 63-63 TSWNN+: Check-in Prediction Based on Deep Learning and Factorization Machine (Chang Su, Ningning Liu, Xianzhong Xie, Shaowen Peng)....Pages 65-79 Deep Learning Based Sentiment Analysis on Product Reviews on Twitter (Aytuğ Onan)....Pages 80-91 A Cluster-Based Machine Learning Model for Large Healthcare Data Analysis (Fatemeh Sharifi, Emad Mohammed, Trafford Crump, Behrouz H. Far)....Pages 92-106 Satire Detection in Turkish News Articles: A Machine Learning Approach (Mansur Alp Toçoğlu, Aytuğ Onan)....Pages 107-117 Front Matter ....Pages 119-119 Committee of the SGTM Neural-Like Structures with Extended Inputs for Predictive Analytics in Insurance (Roman Tkachenko, Ivan Izonin, Michal Greguš ml., Pavlo Tkachenko, Ivanna Dronyuk)....Pages 121-132 Game Analytics on Free to Play (Robert Flunger, Andreas Mladenow, Christine Strauss)....Pages 133-141 A Decentralized File Sharing Framework for Sensitive Data (Onur Demir, Berkay Kocak)....Pages 142-149 The Information System for the Research in Carotid Atherosclerosis (Jiri Blahuta, Tomas Soukup)....Pages 150-161 Front Matter ....Pages 163-163 Big Data Analytics for Financial Crime Typologies (Kirill Plaksiy, Andrey Nikiforov, Natalia Miloslavskaya)....Pages 165-178 Development of a Model for Identifying High-Risk Operations for AML/CFT Purposes (Pavel Y. Leonov, Viktor P. Suyts, Oksana S. Kotelyanets, Nikolai V. Ivanov)....Pages 179-192 Monitoring System for the Housing and Utility Services Based on the Digital Technologies IIoT, Big Data, Data Mining, Edge and Cloud Computing (Vasiliy S. Kireev, Pyotr V. Bochkaryov, Anna I. Guseva, Igor A. Kuznetsov, Stanislav A. Filippov)....Pages 193-205 K-Means Method as a Tool of Big Data Analysis in Risk-Oriented Audit (Pavel Y. Leonov, Viktor P. Suyts, Oksana S. Kotelyanets, Nikolai V. Ivanov)....Pages 206-216 Back Matter ....Pages 217-217