Flexible Query Answering Systems: 15th International Conference, FQAS 2023, Mallorca, Spain, September 5–7, 2023, Proceedings (Lecture Notes in Artificial Intelligence)

This book constitutes the refereed proceedings of the 15th International Conference on Flexible Query Answering Systems, FQAS 2023, held in September 2023 in Palma de Mallorca, Spain. The 24 full papers presented were carefully reviewed and selected from numerous submissions. They are organized in the following topical sections: Flexible Queries over Semantic Systems; Advanced Methods and Applications in Natural Language; Processing (NLP); New Advances in Disinformation Detection; Data and Text Mining; Applying AI to Social Science and Social Science to AI; Artificial Intelligence Law and Regulation. Preface Organization Keynote Speakers Identifying Misinformation Online: Open Issues and Challenges Building an Open Web Index for an Open Web Search Ecosystem Survey of Readiness Levels Methodologies Contents Flexible Queries Over Semantic Systems On Reducing Reasoning and Querying in Natural Logic to Database Querying*-1pc 1 Introduction 2 NaturaLog in a Nutshell 3 Compound Concept Terms 3.1 Proxy Concept Terms 4 Graph Visualisations 5 Inference Rules 6 Deductive Querying via Database Querying 6.1 Specialisation and Generalisation Concept Queries 6.2 Commonality Queries 6.3 Connection Queries 6.4 Exploring Pathways 6.5 Analogies 7 Concluding Summary References Diversifying Top-k Answers in a Query by Example Setting*-1pc 1 Introduction 2 Related Works 2.1 The Query-By-Example Paradigm 2.2 Diversified Search 3 Result Diversification in the QBE Paradigm 3.1 Diversity wrt. a Set of Representative Examples 3.2 Diversification Strategy 4 Illustration 4.1 Div-DCQ 4.2 Experimentations 5 Conclusion and Perspectives References Flexible Classification, Question-Answering and Retrieval with Siamese Neural Networks for Biomedical Texts 1 Introduction 2 Related Works 3 Transformers 3.1 Pre-trained Language Models 3.2 Siamese Models 4 Proposed Models 5 Experiments and Results 5.1 Experiments 5.2 Results 6 Conclusion References The Promise of Query Answering Systems in Sexuality Studies: Current State, Challenges and Limitations*-1pc 1 Introduction 2 Artificial Intelligence in Sexuality Studies 3 Why Sexuality Studies Require AI? 4 The Role of Query Answering Systems in Sexuality Studies 5 Discussion 5.1 Current Challenges 5.2 Limitations 5.3 Social Impact 6 Conclusions References Some Properties of the Left Recursive Form of the Convex Combination Linguistic Aggregator 1 Introduction 2 Background 3 Left Recursive Form of CONV and LOWA 4 Some Properties of CONV–LRF and LOWA–LRF 5 Evaluation of the Orness 6 Use Case: Fuzzy Ontology Building 7 Conclusions and Future Work References Knowledge Graph Enabled Open-Domain Conversational Question Answering 1 Introduction 2 Related Research 3 Converse Architecture 3.1 Graph Generation Module 3.2 Query Processing Module 3.3 Natural Language Generation (NLG) Module 4 Evaluation 5 Future Research 6 Conclusion References Advanced Methods and Applications in Natural Language Processing (NLP) Automatic Generation of Coherent Natural Language Texts 1 Introduction 1.1 The Multifaceted Nature of Natural Language Texts Generation 1.2 Existing Decoding Methods Used for Generation 1.3 Coherence Aspect of the Text Quality 2 Modified Generation Method Based on the Coherence Metric 2.1 Coherence Metric Used 2.2 Generic Modification of Decoding Methods Based on Coherence Metric 2.3 Examples of the Generated Texts 3 Discussion 3.1 Need for Text Quality Estimation Methodologies 3.2 Points for Improvement, Possible Applications, and Further Development of the Method References Interlingual Semantic Validation 1 Introduction 2 Related Work. Assessing Quality in Machine Translation 2.1 Human-Based Evaluation 2.2 Automatic Metrics-Based Evaluation 3 A Model for Semantic Validation 3.1 Description of Graph-Based Representation 3.2 Semantic Validation 4 Quantitative Evaluation 5 Conclusions References How Tasty Is This Dish? Studying User-Recipe Interactions with a Rating Prediction Algorithm and Graph Neural Networks 1 Introduction 2 Related Work 3 Method 3.1 Overview 3.2 Graph Neural Network Model 3.3 Rating Prediction Layer 4 Experiments and Discussion 4.1 Data 4.2 Setup 4.3 Results and Discussion 5 Conclusions and Future Work References ``Let It BEE'': Natural Language Classification of Arthropod Specimens Based on Their Spanish Description 1 Introduction 2 Methods 2.1 Text Cleaning and Preparation 2.2 Classical Machine Learning Classifier Approaches 2.3 Modern Transformer-Based Approaches 3 Results 4 Conclusions and Future Work References New Advances in Disinformation Detection Bot Detection in Twitter: An Overview 1 Introduction 2 Methodology 3 Bots in Twitter 3.1 Bot Definition 3.2 Categorization 3.3 Botnet 4 Bot Detection 4.1 Techniques 4.2 Available Tools 5 Conclusions and Future Challenges References A Fuzzy Approach to Detecting Suspected Disinformation in Videos 1 Introduction 2 Granular Linguistic Model of Phenomena 3 Methodological Proposal 3.1 Conceptualization Framework 3.2 Multimodal Feature Analyzer 3.3 Multimodal Behaviour Interpreter 3.4 Report Generation 4 Case of Study 4.1 Conceptualization Framework 4.2 Multimodal Feature Analyzer 4.3 Multimodal Behaviour Interpreter 4.4 Report Generation 4.5 Example 5 Conclusions References All Trolls Have One Mission: An Entropy Analysis of Political Misinformation Spreaders 1 Introduction 2 Related Works 3 Data 4 Experimentation 4.1 Character-Level Entropy 4.2 Word-Level Entropy 5 Conclusions and Future Work References Data and Text Mining A First Evolutionary Fuzzy Approach for Change Mining with Smart Bands 1 Introduction 2 Background 2.1 Change Mining 2.2 Evolutionary Fuzzy Sytems 3 SDATEP: Smart Bands Drift Analysis Through Emerging Patterns 3.1 Android App 3.2 SDATEP 4 Experimental Framewok 4.1 Data Collected 4.2 Parameters of the Algorithm and Characteristics of the Device 4.3 Quality Measures and Results 5 Results and Analysis of the Study 5.1 Sleep 5.2 Activity 6 Conclusions References Federated Learning in Healthcare with Unsupervised and Semi-Supervised Methods 1 Introduction 2 Methodology 3 The Federated Paradigm 3.1 Data Partitions 3.2 Scale of the Federation 3.3 Governance 4 Federated Learning in Medical Environments 4.1 Data Characteristics in Medical Environments 4.2 Unsupervised Methods 4.3 Semi-Supervised Methods 5 Conclusions and Remaining Challenges References Exploring Hidden Anomalies in UGR'16 Network Dataset with Kitsune 1 Introduction 2 Related Work 3 Methods and Materials 3.1 UGR'16 Network Dataset 3.2 Feature as a Counter 3.3 Kitsune 4 Experimental Setup and Scenarios 4.1 Preconditions 4.2 Experimental Scenarios 4.3 Metrics 5 Results and Discussion 5.1 Results 6 Conclusion and Future Work References An Orthographic Similarity Measure for Graph-Based Text Representations 1 Introduction 2 Background 2.1 Byzantine Book Epigrams 2.2 Graph Databases 2.3 Fuzzy Graphs 3 Graph-Based Text Model 4 Hierarchical Orthographic Similarity Algorithm 4.1 Calculating Word Similarities 4.2 Calculating Verse Similarities 4.3 Calculating Epigram Similarities 5 Experiments 6 Conclusions Bibliography Applying AI to Social Science and Social Science to AI An Unsupervised Approach to Extracting Knowledge from the Relationships Between Blame Attribution on Twitter 1 Introduction 1.1 Blaming Artificial Intelligence 1.2 The Present Study 2 Methods 2.1 Data Collection 2.2 Analytic Strategy 3 Results 3.1 Qualitative 3.2 Quantitative 4 Discussion 4.1 Limitations 5 Conclusion References ``Health Is the Real Wealth'': Unsupervised Approach to Improve Explainability in Health-Based Recommendation Systems*-1pc 1 Introduction 2 Related Work 2.1 Food Recommendation Systems 2.2 Explainability 2.3 User Adhesion 3 Model for the Generator of Explanations 3.1 Data Sources 3.2 Question Generation 3.3 Paragraph Selection 3.4 Question Answering 4 Results 5 Conclusions and Future Work References Are Textual Recommendations Enough? Guiding Physicians Toward the Design of Machine Learning Pipelines Through a Visual Platform 1 Introduction 2 Background 3 Methodology 3.1 Interaction Analysis 3.2 Perceived Usability Evaluation 4 Results 5 Improvements Proposal and Conclusions References Who Is to Blame? Responsibility Attribution in AI Systems vs Human Agents in the Field of Air Crashes 1 Introduction 1.1 Related Work 1.2 This Study 2 Method 2.1 Participants 2.2 Materials 2.3 Procedure 3 Results 3.1 Prospective Judgements 3.2 Retrospective Judgements 4 Discussion 5 Conclusion References Artificial Intelligence Law and Regulation Methodology for Analyzing the Risk of Algorithmic Discrimination from a Legal and Technical Point of View 1 Introduction 2 Algorithmic Discrimination 3 Legal Protection Against Discriminatory Uses of Artificial Intelligence 4 Improvements to Achieve Artificial Intelligence that Respects Fundamental Rights References Data as Wealth, Data Markets and Its Regulation 1 Data as Wealth and Its Possible Onerous Exchange 2 Can Data Be Considered as Goods from a Legal Point of View? 3 Data Market Places or Platforms for Data Exchange 4 Impact of the Digital Market Act on Data Markets 5 Conclusions References ADM in the European Union: An Interoperable Solution 1 Introducing the Interoperability Framework of Regulations (EU) 817 and 818 of 2019 1.1 One of the New Components: The Multiple-Identity Detector 1.2 Accessing the Common Identity Repository for Detecting Multiple Identities 2 The Multiple-Identity Detection Procedure 2.1 The Multiple-Identity Detection Procedure in a Nutshell 2.2 Is There a Right Not to be Subject to the MID Links? 3 Final Remarks Bibliography Author Index Flexible Query Answering Systems is an edited collection of contributed chapters. It focuses on developing computer systems capable of transforming a query into an answer with useful information. The emphasis is on problems associated with high-level intelligent answering systems. The coverage is multidisciplinary with chapters by authors from information science, logic, fuzzy systems, databases, artificial intelligence and knowledge representation. Each contribution represents a theory involving flexibility in query-answering, and each addresses specific answering problems. Flexible Query Answering Systems is a timely contribution for researchers working on high-level query mechanism systems.