Advances in information retrieval : 43rd European Conference on IR Research, ECIR 2021, virtual event, March 28 - April 1, 2021 : proceedings / / Djoerd Hiemstra [and five others], editors

• Pubblicazione: Cham, Switzerland : , : Springer, , [2021]; ©2021
• Descrizione fisica: 1 online resource (xxxiii, 738 pages)
• Disciplina: 025.04
• Soggetto topico: Information retrieval; Data mining; Information storage and retrieval systems
• Persona (resp. second.): HiemstraDjoerd
• Nota di bibliografia: Includes bibliographical references and index.
• Nota di contenuto: Intro -- Preface -- Organization -- Contents - Part I -- Contents - Part II -- Full Papers -- Stay on Topic, Please: Aligning User Comments to the Content of a News Article -- 1 Introduction -- 2 Related Work -- 3 Datasets -- 3.1 Labeling -- 3.2 User Agreement Study -- 4 Methods -- 4.1 BERTAC Model - Joint Modeling of Article and Comments -- 4.2 BA-BC Model - Disjoint Modeling of Article and Comments -- 4.3 Ordinal Classification Loss -- 5 Experimental Setup -- 5.1 Environment -- 5.2 Model Evaluation -- 5.3 Comparison Models -- 6 Results and Discussion -- 6.1 Binary Versus Multiclass ACAP -- 6.2 Models Comparisons on Multiclass ACAP -- 6.3 Weighted Versus Un-Weighted Accuracy -- 6.4 Ordinal Classification Loss -- 7 Conclusion -- References -- An E-Commerce Dataset in French for Multi-modal Product Categorization and Cross-Modal Retrieval -- 1 Introduction -- 2 Related Work -- 3 Challenge Description -- 4 Data Description -- 4.1 Data Characteristics -- 5 Evaluation -- 5.1 Evaluation Metric -- 5.2 Evaluation Phases and Timeline -- 6 Baseline Models -- 7 Short Descriptions of Top Performing Systems -- 7.1 System Performances -- 8 Conclusion -- References -- FedeRank: User Controlled Feedback with Federated Recommender Systems -- 1 Introduction -- 2 Related Work -- 3 Approach -- 3.1 FedeRank -- 4 Experiments -- 4.1 Performance of Federated Learning to Rank -- 4.2 Analysis of Positive Feedback Transmission Ratio -- 4.3 Study on FedeRank Algorithmic Bias -- 5 Conclusion and Future Work -- References -- Active Learning for Entity Alignment -- 1 Introduction -- 2 Problem Setting -- 3 Related Work -- 4 Methodology -- 4.1 Labeling Setting -- 4.2 Dataset Adjustment -- 4.3 Active Learning Heuristics -- 5 Evaluation Framework -- 6 Experiments -- 6.1 Setup -- 6.2 Results -- 7 Conclusion -- References. ; Exploring Classic and Neural Lexical Translation Models for Information Retrieval: Interpretability, Effectiveness, and Efficiency Benefits -- 1 Introduction -- 2 Related Work -- 3 Methods -- 4 Experiments -- 4.1 Setup -- 4.2 Results -- 5 Conclusion -- References -- Coreference Resolution in Research Papers from Multiple Domains -- 1 Introduction -- 2 Related Work -- 2.1 Approaches for Coreference Resolution -- 2.2 Corpora for Coreference Resolution in Research Papers -- 3 Coreference Resolution in Research Papers -- 3.1 Corpus for Coreference Resolution in 10 STM Domains -- 3.2 Transfer Learning for Coreference Resolution -- 3.3 Cross-Domain Research Knowledge Graph Population -- 3.4 Evaluation Procedure of Clustering in KG Population -- 4 Experimental Setup -- 4.1 Automatic Coreference Resolution -- 4.2 Evaluation of KG Population Strategies -- 5 Results and Discussion -- 5.1 Automatic Coreference Resolution -- 5.2 Cross-Domain Research KG -- 6 Conclusions -- References -- How Do Simple Transformations of Text and Image Features Impact Cosine-Based Semantic Match? -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Feature Transforms -- 3.2 Complementary Experiments -- 4 Tasks and Data -- 4.1 Task Descriptions -- 4.2 Datasets -- 5 Experimental Setup -- 5.1 Feature Vectors (Embeddings) -- 5.2 Training Setup and Implementation -- 6 Results -- 6.1 Real-World Tasks -- 6.2 Synthetic Data -- 7 Conclusions and Future Work -- References -- An Enhanced Evaluation Framework for Query Performance Prediction -- 1 Introduction -- 2 Related Work -- 3 Experimental Analysis -- 3.1 Experimental Setup -- 3.2 Traditional QPP Evaluation Using Correlations -- 3.3 ANOVA Modeling and Analysis of QPP -- 3.4 ANOVA Modeling of Multiple Queries and Interactions -- 4 Conclusion -- References -- Open-Domain Conversational Search Assistant with Transformers -- 1 Introduction. ; 2 Related Work -- 3 Transformers-Based Conversational Search Assistant -- 3.1 Conversational Query Rewriting Transformer -- 3.2 Passage Re-Ranking Transformer -- 3.3 Abstractive Search-Answer Generation Transformer -- 4 Evaluation -- 4.1 Datasets and Protocol -- 4.2 Implementation -- 4.3 Results and Discussion -- 5 Conclusions -- References -- Complement Lexical Retrieval Model with Semantic Residual Embeddings -- 1 Introduction -- 2 Related Work -- 3 Proposed Method -- 3.1 Lexical Retrieval Model -- 3.2 Embedding Retrieval Model -- 3.3 Residual-Based Learning -- 3.4 Retrieval with CLEAR -- 4 Experimental Methodology -- 5 Results and Discussion -- 5.1 Retrieval Accuracy of CLEAR -- 5.2 Impacts of clear on Reranking -- 5.3 Case Study: The Goods and the New Challenges -- 6 Conclusion -- References -- Classifying Scientific Publications with BERT - Is Self-attention a Feature Selection Method? -- 1 Introduction -- 2 Related Work -- 3 Fine-Tuning Language Models for Text Classification -- 4 Exploring Self-attention Heads -- 5 Feature Selection -- 5.1 Domain Knowledge -- 5.2 Feature Evaluation -- 6 Conclusions -- References -- Valuation of Startups: A Machine Learning Perspective -- 1 Introduction -- 2 Related Work -- 3 Data Analysis and Problem Formulation -- 3.1 Source and Target Data: Crunchbase -- 3.2 TargetLAB Data: Companies House -- 3.3 Summary of Dataset -- 4 Approach -- 4.1 Unsupervised Domain Adaptation -- 4.2 Supervised Domain Adaptation -- 4.3 Semi-supervised Domain Adaptation -- 4.4 Features -- 5 Experiments -- 5.1 Baselines -- 5.2 Experimental Setup and Metrics -- 5.3 Results -- 5.4 Feature Group Contributions -- 6 Conclusion -- References -- Disparate Impact in Item Recommendation: A Case of Geographic Imbalance -- 1 Introduction -- 2 Related Work -- 3 Preliminaries -- 3.1 Recommendation Scenario -- 3.2 Metrics. ; 3.3 Recommendation Algorithms -- 3.4 Datasets -- 4 Disparate Impact Assessment -- 4.1 Experimental Setting -- 4.2 Characterizing User Behavior -- 4.3 Assessing Effectiveness and Disparate Impact -- 5 Mitigating Disparate Impact -- 5.1 Algorithm -- 5.2 Impact of Mitigation -- 6 Conclusions and Future Work -- References -- You Get What You Chat: Using Conversations to Personalize Search-Based Recommendations -- 1 Introduction -- 2 Computational Model and Re-ranking -- 2.1 Re-ranking Methods -- 3 Domain-Specific Vocabulary Weighting -- 4 Entity Expansion -- 5 Data Collection -- 6 Experimental Studies -- 6.1 Setup -- 6.2 User Models: None vs. Chats vs. Questionnaires (RQ0 and RQ1) -- 6.3 Domain Vocabularies (RQ2) -- 6.4 Entity Expansion (RQ3) -- 7 Related Work -- 8 Conclusion -- References -- Joint Autoregressive and Graph Models for Software and Developer Social Networks -- 1 Introduction -- 2 Related Work -- 3 Dataset -- 4 Notation and Preliminaries -- 5 Bug Urgency Ranking -- 6 Developer Recommendation -- 7 Experiments and Results -- 7.1 Bug Urgency Ranking -- 7.2 Developer Recommendation -- 8 Discussion and Conclusion -- References -- Mitigating the Position Bias of Transformer Models in Passage Re-ranking -- 1 Introduction -- 2 Background -- 2.1 Transformer -- 2.2 BERTCAT Ranking Model -- 2.3 BERTDOT Ranking Model -- 2.4 TK Ranking Model -- 3 Experiment Design -- 4 Dataset Analysis and Debiasing -- 4.1 Dataset Analysis -- 4.2 Debiasing the Passage Datasets -- 5 Transformer Bias Analysis -- 6 Retrieval Results -- 7 Related Work -- 8 Conclusion -- References -- Exploding TV Sets and Disappointing Laptops: Suggesting Interesting Content in News Archives Based on Surprise Estimation -- 1 Introduction -- 2 Related Works -- 3 Proposed Approaches -- 3.1 Input Data -- 3.2 Topic-Based Mutually Reinforced Random Walk. ; 3.3 Topic Pair-Based Mutually Reinforced Random Walk -- 4 Experimental Settings -- 4.1 Temporal Document Collection -- 4.2 Preprocessing -- 4.3 Baselines -- 4.4 Data Annotation -- 5 Experimental Results -- 5.1 Main Results -- 5.2 Case Studies -- 6 Conclusions -- References -- Label Definitions Augmented Interaction Model for Legal Charge Prediction -- 1 Introduction -- 2 Related Works -- 3 The Proposed Model -- 3.1 Problem Formulation -- 3.2 Framework -- 3.3 Fact Description Encoder -- 3.4 Charge Definitions Encoder -- 3.5 Two Auxiliary Fact Representations from Charge Definitions -- 3.6 The Output -- 4 Experiments -- 4.1 Datasets -- 4.2 Training Setup -- 4.3 Baselines -- 4.4 Results -- 4.5 Ablation Test -- 4.6 Intra-class Variance of Different Fact Representations -- 4.7 Case Study -- 5 Conclusion -- References -- A Study of Distributed Representations for Figures of Research Articles -- 1 Introduction -- 2 Related Work -- 3 Figure Embeddings -- 4 Weak Supervision for Figure Embeddings -- 5 Empirical Study -- 5.1 Experimental Setup -- 5.2 Experimental Results -- 6 Conclusions -- References -- Answer Sentence Selection Using Local and Global Context in Transformer Models -- 1 Introduction -- 2 Related Work -- 3 Transformer Models for Answer Sentence Selecting -- 3.1 AS2 Definition -- 3.2 Selecting Sentences with a Transformer Model -- 4 Contextual Transformer for AS2 -- 4.1 Local Context -- 4.2 Global Context -- 4.3 Combined Context -- 5 Empirical Assessment -- 5.1 Corpora -- 5.2 Models -- 5.3 Results -- 6 Conclusion -- References -- An Argument Extraction Decoder in Open Information Extraction -- 1 Introduction -- 2 The Open IE Framework with Our Proposed Decoder -- 2.1 Predicate Extractions -- 2.2 The Predicate-Specific Embedding Layer -- 2.3 Argument Extraction with the Proposed Decoder -- 2.4 Training -- 3 Experiments -- 3.1 Data -- 3.2 Settings. ; 3.3 Baselines.
• Titolo autorizzato: Advances in Information Retrieval
• ISBN: 3-030-72113-2
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Lingua di pubblicazione: Inglese
• Record Nr.: 9910485030603321
• Lo trovi qui: Univ. Federico II
• Serie: Lecture notes in computer science ; ; 12656.