[Communications in Computer and Information Science] Machine Learning and Knowledge Discovery in Databases Volume 1168 (International Workshops of ECML PKDD 2019, WÃ1⁄4rzburg, Germany, September 16–20, 2019, Proceedings, Part II) ||

This two-volume set constitutes the refereed proceedings of the workshops which complemented the 19th Joint European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD, held in Würzburg, Germany, in September 2019. The 70 full papers and 46 short papers presented in the two-volume set were carefully reviewed and selected from 200 submissions. The two volumes (CCIS 1167 and CCIS 1168) present the papers that have been accepted for the following workshops: Workshop on Automating Data Science, ADS 2019; Workshop on Advances in Interpretable Machine Learning and Artificial Intelligence and eXplainable Knowledge Discovery in Data Mining, AIMLAI-XKDD 2019; Workshop on Decentralized Machine Learning at the Edge, DMLE 2019; Workshop on Advances in Managing and Mining Large Evolving Graphs, LEG 2019; Workshop on Data and Machine Learning Advances with Multiple Views; Workshop on New Trends in Representation Learning with Knowledge Graphs; Workshop on Data Science for Social Good, SoGood 2019; Workshop on Knowledge Discovery and User Modelling for Smart Cities, UMCIT 2019; Workshop on Data Integration and Applications Workshop, DINA 2019; Workshop on Machine Learning for Cybersecurity, MLCS 2019; Workshop on Sports Analytics: Machine Learning and Data Mining for Sports Analytics, MLSA 2019; Workshop on Categorising Different Types of Online Harassment Languages in Social Media; Workshop on IoT Stream for Data Driven Predictive Maintenance, IoTStream 2019; Workshop on Machine Learning and Music, MML 2019; Workshop on Large-Scale Biomedical Semantic Indexing and Question Answering, BioASQ 2019. Preface Organization Contents – Part II Contents – Part I Second International Workshop on Knowledge Discovery and User Modeling for Smart Cities (UMCit) District Heating Substation Behaviour Modelling for Annotating the Performance 1 Introduction 2 Methods and Techniques 2.1 Sequential Pattern Mining 2.2 Clustering Analysis 2.3 Distance Measure 3 Proposed Method 4 Results and Discussion 5 Conclusion and Future Work References Modeling Evolving User Behavior via Sequential Clustering 1 Introduction 2 Modeling Evolving User Behavior via Sequential Clustering 2.1 Sequential Partitioning Algorithm 2.2 Partitioning Algorithms 2.3 Dynamic Time Warping Algorithm 3 Case Study: Modeling Household Electricity Consumption Behavior 3.1 Case Description 3.2 Data and Experiments 3.3 Results and Discussion 4 Conclusions and Future Work References Recognizing User's Activity and Transport Mode Detection: Maintaining Low-Power Consumption 1 Introduction 2 Related Work 3 Data Collection 4 Data Preprocessing 5 Feature Extraction 6 Classification Models 6.1 Accelerometer Features 7 Evaluation and Results 7.1 Effect of Filtering Raw Accelerometer Data on Classification Accuracy 7.2 Effect of Location Data on Classification Accuracy 8 Conclusion References Can Twitter Help to Predict Outcome of 2019 Indian General Election: A Deep Learning Based Study 1 Introduction 2 Literature Review 3 Methodology 3.1 Dataset Collection 3.2 Hashtags Based Tweets Segregation 3.3 Twitter Sentiment Classification 3.4 Opinion Analysis Corresponding to Different States 4 Conclusion References Towards Sensing and Sharing Auditory Context Information Using Wearable Device 1 Introduction 2 Wearable Ambient Sound Sensing System 2.1 Wearable Device 2.2 Auditory Sensing Data 3 Extracting Context Information 3.1 Segmentation of Multi-dimensional Time-Series Data 4 Discussion and Future Work 5 Conclusion References Workshop on Data Integration and Applications (DINA) Noise Reduction in Distant Supervision for Relation Extraction Using Probabilistic Soft Logic 1 Introduction 2 Related Work 3 Probabilistic Soft Logic 4 HL-MRF Model for Noise Reduction 4.1 Prior Model 4.2 Consistency Between Predictions of NER Systems 4.3 Sentence Structure Analysis 4.4 Context-Based Constraints 4.5 Semantic Similarity in Noise Reduction 5 Experimental Evaluation 5.1 Experimental Setup: Data and Models 5.2 Experimental Setup: Benchmark Methods 5.3 Experimental Results 6 Conclusions and Outlook References Privacy-Preserving Record Linkage to Identify Fragmented Electronic Medical Records in the All of Us Research Program 1 Introduction 1.1 The All of Us Research Program 1.2 Data Fragmentation Across Institutions 1.3 Prior Use of Privacy-Preserving Record Linkage 2 Methods 3 Results 3.1 Patient with Care Fragmentation 3.2 Data Quality Issues 3.3 Geographic Analysis to Characterize “Snowbirds” 4 Discussion References Data Integration for the Development of a Seismic Loss Prediction Model for Residential Buildings in New Zealand 1 Background 1.1 The Christchurch Earthquake Sequence 1.2 Seismic Insurance Following the Canterbury Earthquake Sequence 1.3 The Earthquake Commission 1.4 EQC's Catastrophe Loss Models 1.5 Earthquake Commission Amendment Bill 2 Developing a Loss Prediction Model Using EQC's Residential Claim Database 2.1 Exploration of the Database 2.2 Merging of Multiple Databases 2.3 Challenges and Lessons Learned 3 Future Model Development Using Machine Learning 4 Conclusion References Linking IT Product Records 1 Introduction 2 Related Work 3 Hybrid Similarity Measure 4 Experimental Evaluation 5 Conclusions and Future Work References Pharos: Query-Driven Schema Inference for the Semantic Web 1 Introduction 2 Approach 3 Related Work 4 Contribution 4.1 Schema Inference 4.2 Ontology 5 Prototype 6 Evaluation 7 Future Work 8 Summary References Informativeness-Based Active Learning for Entity Resolution 1 Introduction 2 Related Work 3 Problem Definition 4 Informativeness-Aware Active Learning 4.1 Initial Selection 4.2 Informativeness of Similarity Vectors 4.3 Training Data Selection 4.4 Complexity Analysis 5 Experiments and Results 5.1 Parameter Evaluation 5.2 Comparison with Existing Approaches 6 Conclusions and Future Work References Encoding Hierarchical Classification Codes for Privacy-Preserving Record Linkage Using Bloom Filters 1 Introduction 2 Methods 2.1 Bloom Filters 2.2 Positional BFs (PBFs) 2.3 Hierarchy Preserving Bloom Filters (HPBFs) 2.4 Data 2.5 Evaluation Methods 3 Results 3.1 Privacy-Preserving Record Linkage (PPRL) 4 Conclusion References Machine Learning for Cybersecurity (MLCS) Are Network Attacks Outliers? A Study of Space Representations and Unsupervised Algorithms 1 Introduction 2 Problem Spaces in NTA 3 Outlier Detection Algorithms 4 Dataset 5 Experiments 6 Outlier Detection Evaluation Indices 7 Results and Discussion 7.1 Are Network Attacks Outliers? 7.2 What Are the Best Feature Vectors for the Task? 7.3 Can We Improve Vectors and Use Them in Real Detection? 8 Conclusions References Auto Semi-supervised Outlier Detection for Malicious Authentication Events 1 Introduction 2 Related Work 3 Methodology 3.1 Phase 1 3.2 Phase 2 4 Experiments and Evaluation 4.1 Dataset 4.2 Experiment Environment 4.3 Experimental Settings 4.4 Evaluation 5 Conclusion and Future Work References Defense-VAE: A Fast and Accurate Defense Against Adversarial Attacks 1 Introduction 2 Defense-VAE: The Proposed Algorithm 2.1 Variational Auto-Encoder 2.2 Defense-VAE 3 Related Work 4 Experiments 4.1 Results on White-Box Attacks 4.2 Robustness Under Untrained Attacks 4.3 Results on Black-Box Attacks 4.4 Why Is Defense-VAE so Effective? 4.5 Defense Speed 5 Conclusion A Network Architectures B Experiments on CelebA and CIFAR-10 References Analyzing and Storing Network Intrusion Detection Data Using Bayesian Coresets: A Preliminary Study in Offline and Streaming Settings 1 Introduction 2 Background 2.1 Notation 2.2 Bayesian Machine Learning 2.3 Bayesian Coresets 2.4 Alternative Approaches to BCH for Bayesian Learning 3 Network Security 4 Experimental Setup 5 Simulation 1: BCH Applied to Network Intrusion Detection Data 6 Simulation 2: BCH in a Streaming Environment 7 Conclusion and Future Work References 6th Workshop on Sports Analytics: Machine Learning and Data Mining for Sports Analytics (MLSA) Analyzing Soccer Players' Skill Ratings Over Time Using Tensor-Based Methods 1 Introduction 2 Tensors and the SoFIFA Data 2.1 Tensors 2.2 Data Description 2.3 Data Challenges 3 Exposing Hidden Structure with CPD 4 Predicting Skill Ratings Using Tucker Decomposition 4.1 Tucker Decomposition Theory 4.2 Predicting Skill Ratings 4.3 Experiments 5 Conclusions References Exploring Successful Team Tactics in Soccer Tracking Data 1 Introduction 1.1 Tactical Analysis in Practice 1.2 Positional Tracking Data 1.3 Aims 2 Methodological Approach 2.1 Key Events 2.2 Feature Construction 2.3 Discovering Patterns 3 Experiment 3.1 Data 3.2 Subgroup Discovery 3.3 Subgroups 4 Discussion 4.1 Conclusions References Soccer Team Vectors 1 Introduction 2 Related Work 3 Soccer Team Vectors 3.1 Overview 3.2 Problem Definition 3.3 Algorithm 4 Experiments 4.1 Dataset and Experimental Setup 4.2 Similarity Search 4.3 Ranking Soccer Teams 4.4 Team Market Value Estimation 5 Conclusion References Tactical Analyses in Professional Tennis 1 Introduction 2 Materials 2.1 Data 2.2 Data Preparation 3 Feature Engineering 3.1 Point Characteristics (4 Features) 3.2 Match Situation (13 Features) 3.3 Stroke Characteristics (27 Features) 3.4 Rally Features (61 Features) 4 Experiments 4.1 Methods 4.2 Serve Characteristics 5 Conclusion References Difficulty Classification of Mountainbike Downhill Trails Utilizing Deep Neural Networks 1 Introduction 1.1 Related Work 2 The Dataset 2.1 Collecting and Labeling Data 2.2 Input Data Representation 3 Classification Through a 2D Convolutional Neural Network 3.1 Experiments 4 Conclusion References First Workshop on Categorizing Different Types of Online Harassment Languages in Social Media Categorizing Online Harassment on Twitter 1 Introduction 2 Related Work 3 Methodology 3.1 Data 3.2 Algorithms 3.3 Word Embeddings 4 Experiments 4.1 Pre-processing 4.2 Task A – Binary Classification 4.3 Task B – Multi-class Classification 4.4 Classification with Word Embeddings 5 Results and Discussion 5.1 Classification with TF-IDF Vectors 5.2 Classification with Word Embeddings 6 Conclusion and Future Works References Learning to Detect Online Harassment on Twitter with the Transformer 1 Introduction 2 Related Work 3 Proposal 3.1 Applying Self-attention 3.2 Baselines 4 Experiments 4.1 Data 4.2 Traning 5 Results and Discussion 6 Conclusions References Detection of Harassment on Twitter with Deep Learning Techniques 1 Introduction 2 Related Work 3 Method 4 Experiments and Results 4.1 Dataset and Evaluation 4.2 Data Processing 4.3 Harassment Detection 5 Conclusions References Gradient Boosting Machine and LSTM Network for Online Harassment Detection and Categorization in Social Media 1 Introduction 2 Dataset Description 3 Feature Engineering 3.1 Text Pre-processing Pipeline 3.2 Feature Extraction 3.3 Visualizing Tweets as a Similarity Network 4 Proposed Approaches 4.1 Task A: Online Harassment Detection 4.2 Task B: Categorizing Types of Online Harassment 5 Experimental Results 6 Conclusion References Attention-Based Method for Categorizing Different Types of Online Harassment Language 1 Introduction 2 Related Work 3 Dataset Description 4 Proposed Methodology 4.1 Data Augmentation 4.2 Text Processing 4.3 RNN Model and Attention Mechanism 4.4 Model Architecture 5 Experiments 5.1 Training Models 5.2 Evaluation and Results 6 Conclusion - Future Work References IoT Stream for Data Driven Predictive Maintenance SPICE: Streaming PCA Fault Identification and Classification Engine in Predictive Maintenance 1 Introduction 2 Related Work 3 Methods 4 Materials 5 Experiments and Discussion 6 Conclusion References Event-Based Predictive Maintenance on Top of Sensor Data in a Real Industry 4.0 Case Study 1 Introduction 2 The Case Study 3 A Supervised Learning Technique Based on a Timeseries Discretization Methodology 3.1 Data Pre-processing and MP-Based Timeseries Analytics 3.2 The MP-based Algorithm to Extract Artificial Events 3.3 Supervised Event-Based PdM Approach Using Artificial Events 3.4 Event-Based PdM Solution Details 3.5 Experiments and Results 4 An Unsupervised Learning Technique 5 Discussion 6 Related Work 7 Conclusions and Future Work References Forecasting of Product Quality Through Anomaly Detection 1 Introduction 2 Related Work 3 Data 4 Initial Analyses of Data 4.1 Embedding the Data in 2D Space 4.2 Clustering the Data 4.3 Applying Classifiers 4.4 One-Class Classification 4.5 Applying Long Short-Term Memory Network Model 5 Conclusion References Data Preprocessing and Dynamic Ensemble Selection for Imbalanced Data Stream Classification 1 Introduction 2 Dynamic Ensemble Selection and Data Preprocessing 3 The Proposed Framework 4 Experimental Evaluation 4.1 Experimental Setup 4.2 Lessons Learned 5 Conclusions References A Study on Imbalanced Data Streams 1 Introduction 2 Related Work 3 Problem Definition 4 Proposed Method 4.1 Under-Sampling 4.2 Over-Sampling 5 Experimental Results 6 Conclusions and Future Work References Mining Human Mobility Data to Discover Locations and Habits 1 Introduction 2 Related Work 3 Problem Statement and Methodology 3.1 User Stay Points Detection 3.2 Meaningful Locations 3.3 Identification of Habits 3.4 Gaussian Mixture Model to Classify the Different Habits 4 Experiments and Results 4.1 Datasets 4.2 Clustering Results 4.3 Habits Results 5 Conclusions and Future Work References Imbalanced Data Stream Classification Using Hybrid Data Preprocessing 1 Introduction 2 Related Works 3 The Deterministic Sampling Classifier 4 Experimental Evaluation 5 Conclusions and Future Directions References A Machine Learning-Based Approach for Predicting Tool Wear in Industrial Milling Processes 1 Introduction 2 Related Work 3 Model Selection 3.1 Gradient Boosting Machine 3.2 Temporal Convolutional Network 4 Experiments 4.1 Data and Experimental Setup 4.2 Preprocessing 4.3 Feature Selection and Hyperparameter Tuning 4.4 Results 5 Conclusion and Future Work References 12th International Workshop on Machine Learning and Music (MML 2019) Cross-version Singing Voice Detection in Opera Recordings: Challenges for Supervised Learning 1 Introduction 2 Deep-Learning Methods 3 Dataset 4 Experiments References Neural Symbolic Music Genre Transfer Insights 1 Introduction 2 Related Work 3 Methodology 3.1 Dataset 3.2 Architecture 3.3 Metrics 3.4 Genre Attribution 4 Experiments and Results 4.1 Genre Transfer 4.2 Attribution 5 Conclusion References Familiar Feelings: Listener-Rated Familiarity in Music Emotion Recognition 1 Introduction 2 Related Work 2.1 Dataset 2.2 Labeling 2.3 Algorithms and Analysis 3 Results 4 Conclusion and Future Work References Rhythm, Chord and Melody Generation for Lead Sheets Using Recurrent Neural Networks 1 Lead Sheets 2 The Wikifonia Dataset 2.1 Preprocessing 2.2 Data Encoding and Features 3 Recurrent Neural Network Design 3.1 Optimization Details 4 Experiments 5 Conclusion A Mode Mapping for Chords B Rhythm types References Bacher than Bach? On Musicologically Informed AI-Based Bach Chorale Harmonization 1 Introduction 2 Musicologically Informed Harmonization 2.1 Data Processing and Augmentation 2.2 Network Architecture 2.3 Beam Search 3 Generation Results 4 Evaluation References Adaptively Learning to Recognize Symbols in Handwritten Early Music 1 Introduction 2 Data Structure 3 Incremental Learning 4 Data and Results 5 Conclusion References Feature-Based Classification of Electric Guitar Types 1 Introduction 2 Methodology 2.1 Electric Guitars Audio Recordings Dataset 2.2 New Approach: Feature Analysis and Classification Algorithm 3 Classification Experiments 3.1 Binary Classification 3.2 Multiclass Classification 4 Conclusion References RecurSIA-RRT: Recursive Translatable Point-Set Pattern Discovery with Removal of Redundant Translators 1 Introduction 2 The RecurSIA Algorithm 3 The RRT Algorithm 4 Evaluation 5 Conclusion References Bow Gesture Classification to Identify Three Different Expertise Levels: A Machine Learning Approach 1 Introduction 1.1 Motivation 1.2 Gesture Recognition in Musical Context 2 Methods and Materials 2.1 Music Score 2.2 Recordings and Synchronization 2.3 OpenFrameworks Visualization 2.4 Machine Learning Model 3 Results 4 Discussion and Conclusions 4.1 Future Work References Symbolic Music Classification Based on Multiple Sequential Patterns 1 Introduction 2 Pattern-Based Classification 2.1 Pattern Representation 2.2 Pattern Discovery 2.3 Pattern Ranking and Selection 2.4 Pattern-Based Prediction 3 Results 4 Conclusions References OPTISIA: An Evolutionary Approach to Parameter Optimisation in a Family of Point-Set Pattern-Discovery Algorithms 1 Introduction 2 Previous Work on Parameter Tuning with Genetic Algorithms 3 OMNISIA 4 OPTISIA: An Evolutionary Approach to Parameter Optimisation in OMNISIA 5 Evaluation 6 Conclusion and Suggestions for Future Work References Predicting Dynamics in Violin Pieces with Features from Melodic Motifs 1 Introduction 1.1 Related Work 2 Materials and Methods 2.1 Materials 2.2 Methods 3 Results and Discussion 3.1 Results 3.2 Discussion References Sequence Generation Using Unwords 1 Introduction 2 Methods 2.1 Sampling into Templates 2.2 Information Peaks 2.3 Unwords 2.4 Generation Using Unwords 3 Results 4 Conclusions References A Machine Learning Approach to Study Expressive Performance Deviations in Classical Guitar 1 Introduction 2 Materials and Methods 2.1 Framework 3 Results 4 Conclusion References Enhanced De-Essing via Neural Networks 1 Introduction 2 Conventional De-Essing 3 Data Preparation 4 Prototype Architecture 5 Evaluation 6 Summary References Representation, Exploration and Recommendation of Playlists 1 Introduction 2 Seq2Seq Learning 3 Embedding Models 4 Experimental Setup 4.1 Data: Source and Filtering 4.2 Data Labeling: Genre Assignment 4.3 Training 5 Evaluation Tasks 6 Results 7 Conclusions References Large-Scale Biomedical Semantic Indexing and Question Answering (BioASQ) Results of the Seventh Edition of the BioASQ Challenge 1 Introduction 2 Overview of the Tasks 2.1 Large-Scale Semantic Indexing - 7a 2.2 Biomedical Semantic QA - 7b 3 Overview of Participants 3.1 Task 7a 3.2 Task 7b 4 Results 4.1 Task 7a 4.2 Task 7b 5 Conclusions References Selected Approaches Ranking Contextual Term for the BioASQ Multi-label Classification (Task6a and 7a) 1 Introduction 2 Related Work 3 Methodology 3.1 Multi-label Classification 3.2 Label Ranking 3.3 Multi-label Ranking 4 Results 4.1 System for Multi-label Classification in Task6a 4.2 System for Multi-label Ranking in Task7a 5 Conclusion 6 Discussion References Convolutional Neural Network for Automatic MeSH Indexing 1 Introduction 2 Related Work 3 Methods 3.1 Title and Abstract Embeddings 3.2 Journal Embedding 3.3 Year Encoding 3.4 Hidden and Classification Layers 3.5 Optimization 4 Experiments 4.1 Dataset 4.2 Evaluation Metric 4.3 Configuration 4.4 Evaluation Results 5 Ablation Study 6 Discussion 7 Conclusion References A Mixed Information Source Approach for Biomedical Question Answering: MindLab at BioASQ 7B 1 Introduction 2 The MindLab System at a Glance 2.1 Document Re-Ranking 2.2 Snippet Retrieval 3 Model Performance Tuning 3.1 Document Retrieval 3.2 Snippet Retrieval 4 Results at BioASQ 2019 and Discussion 4.1 Document Retrieval 4.2 Snippet Retrieval 5 Conclusion References AUEB at BioASQ 7: Document and Snippet Retrieval 1 Introduction 2 Document Retrieval Models 2.1 Term-PACRR 2.2 BERT Based Document Retrieval 3 Snippet Retrieval Models 3.1 BCNN 3.2 PDRMM 4 Joint Document and Snippet Retrieval Models 4.1 JPDRMM 5 Overall System Architecture 6 Experiments 6.1 Data and Experimental Setup 6.2 Official Submissions 6.3 Results 7 Related Work 7.1 Document Retrieval 7.2 Snippet Extraction 8 Discussion and Future Work References Classification Betters Regression in Query-Based Multi-document Summarisation Techniques for Question Answering 1 Introduction 2 Related Work 3 Classification vs. Regression Experiments 4 Deep Learning Models 5 Reinforcement Learning 6 Evaluation Correlation Analysis 7 Submitted Runs 8 Conclusions References Structured Summarization of Academic Publications 1 Introduction 2 Related Work 2.1 Summarization Methods 2.2 Summarization Datasets 3 Summarizing Academic Papers 3.1 Flat Abstract Summarization 3.2 SUSIE 4 PMC Structured Abstracts 5 Experiments 5.1 Experimental Setup 5.2 Results 6 Conclusion A Appendix A.1 Reference Summary A.2 Flat P-Gen + Coverage A.3 SUSIE P-Gen + Coverage References How to Pre-train Your Model? Comparison of Different Pre-training Models for Biomedical Question Answering 1 Introduction 2 State of the Art 2.1 Domain Adaptation 2.2 Deep Learning and Domain Adaptation in BIOASQ 3 Question Answering Tasks and Models 3.1 Tasks 3.2 Reading Comprehension - DRQA Model 3.3 Open QA - PSPR Model 3.4 Domain Adaptation for BIOASQ Task 4 Experiments and Results 4.1 Importance of Pre-training and Fine-Tuning 4.2 Experiments with the Two QA Modellings: DRQA and PSPR 4.3 Experiments with Longer Contexts (Modified BIOASQ Data) 5 Conclusion References Yes/No Question Answering in BioASQ 2019 1 Introduction 2 Methods 2.1 ELMo Embeddings 2.2 ELMo Embeddings and Sentiment 2.3 Similarity Matrix 3 Experimental Setup and Results 4 Related Work 5 Conclusions References Semantically Corroborating Neural Attention for Biomedical Question Answering 1 Introduction 2 Literature Review 3 Biomedical QA Methodology 3.1 Factoid QA 3.2 Extractive Answer Selection Module 3.3 Yes/No Answering Model 4 Experimental Analysis 4.1 Factoid QA 4.2 Summary and Yes/No QA 5 Conclusion and Future Insights References Measuring Domain Portability and Error Propagation in Biomedical QA 1 Introduction 2 Related Work 3 BERT Model 4 Systems Overview 4.1 No In-Domain Training 4.2 BioASQ Fine-Tuning 4.3 Snippet Retrieval 4.4 Yes/No and List Question Types 5 Results 5.1 Domain Portability 5.2 Error Propagation 6 Conclusion References UNCC Biomedical Semantic Question Answering Systems. BioASQ: Task-7B, Phase-B 1 Introduction 2 Related Work 2.1 BioASQ 2.2 A Minimum Background on BERT 2.3 Comparison of BERT and Bio-BERT 3 Experiments: Factoid Question Answering Task 3.1 Setup 3.2 Training and Error Analysis 4 Our Systems and Their Performance on Factoid Questions 4.1 LAT Feature Considered and Its Impact (Slightly Negative) 4.2 Impact of Training Using BioASQ Data (Slightly Negative) 4.3 Impact of Using Context from URLs (Negative) 5 Performance on Yes/No and List Questions 5.1 Entailment Improves Yes/No Accuracy 5.2 For List-Type the URLs Have Negative Impact 6 Summary of Our Results 6.1 Factoid Questions 6.2 List Questions 6.3 Yes/No Questions 7 Discussion, Future Experiments, and Conclusions References Transformer Models for Question Answering at BioASQ 2019 1 Introduction 2 Dataset 3 Models 3.1 Models for YesNo Questions 3.2 Models for List Questions 3.3 Model for Factoid Questions 3.4 Model for Summary and Ideal Questions 4 Results 4.1 Ideal Answers 5 Conclusions References Pre-trained Language Model for Biomedical Question Answering 1 Introduction 2 Methods 2.1 BioBERT 2.2 Task-Specific Layer 2.3 Pre-processing 2.4 Post-processing 3 Experimental Setup 3.1 Dataset 3.2 Training 4 Results and Discussion 4.1 Results on BioASQ 7b 4.2 Validating on the BioASQ 6b Dataset 5 Conclusion References Author Index This two-volume set constitutes the refereed proceedings of the workshops which complemented the 19th Joint European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD, held in W赲zburg, Germany, in September 2019. The 70 full papers and 46 short papers presented in the two-volume set were carefully reviewed and selected from 200 submissions. The two volumes (CCIS 1167 and CCIS 1168) present the papers that have been accepted for the following workshops: Workshop on Automating Data Science, ADS 2019; Workshop on Advances in Interpretable Machine Learning and Artificial Intelligence and eXplainable Knowledge Discovery in Data Mining, AIMLAI-XKDD 2019; Workshop on Decentralized Machine Learning at the Edge, DMLE 2019; Workshop on Advances in Managing and Mining Large Evolving Graphs, LEG 2019; Workshop on Data and Machine Learning Advances with Multiple Views; Workshop on New Trends in Representation Learning with Knowledge Graphs; Workshop on Data Science for Social Good, SoGood 2019; Workshop on Knowledge Discovery and User Modelling for Smart Cities, UMCIT 2019; Workshop on Data Integration and Applications Workshop, DINA 2019; Workshop on Machine Learning for Cybersecurity, MLCS 2019; Workshop on Sports Analytics: Machine Learning and Data Mining for Sports Analytics, MLSA 2019; Workshop on Categorising Different Types of Online Harassment Languages in Social Media; Workshop on IoT Stream for Data Driven Predictive Maintenance, IoTStream 2019; Workshop on Machine Learning and Music, MML 2019; Workshop on Large-Scale Biomedical Semantic Indexing and Question Answering, BioASQ 2019. This two-volume set constitutes the refereed proceedings of the workshops which complemented the 19th Joint European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD, held in Wurzburg, Germany, in September 2019.