Advanced Intelligent Computing Technology and Applications : 20th International Conference, ICIC 2024, Tianjin, China, August 5–8, 2024, Proceedings, Part III / / edited by De-Shuang Huang, Zhanjun Si, Qinhu Zhang

• Record Nr.: UNINA9910878980703321
• Autore: Huang De-Shuang
• Pubbl/distr/stampa: Singapore : , : Springer Nature Singapore : , : Imprint : Springer, , 2024
• ISBN: 9789819756698; 9789819756681
• Edizione: [1st ed. 2024.]
• Descrizione fisica: 1 online resource (544 pages)
• Collana: Lecture Notes in Artificial Intelligence, , 2945-9141 ; ; 14877
• Altri autori (Persone): SiZhanjun; ZhangQinhu
• Disciplina: 006.3
• Soggetti: Artificial intelligence; Computers; Computer networks; Data mining; Image processing - Digital techniques; Computer vision; Software engineering; Artificial Intelligence; Computing Milieux; Computer Communication Networks; Data Mining and Knowledge Discovery; Computer Imaging, Vision, Pattern Recognition and Graphics; Software Engineering
• Lingua di pubblicazione: Inglese
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Nota di contenuto: Intro -- Preface -- Organization -- Contents - Part III -- Natural Language Processing and Computational Linguistics -- STAR: Syntax- and Topic-Aware Role Dialogue Summarization -- 1 Introduction -- 2 Related Work -- 3 Methodology -- 3.1 Task Definition -- 3.2 Role Prompts -- 3.3 Utterance Mask -- 3.4 Syntax- and Topic-Aware Role Model -- 4 Experiments -- 4.1 Basic Settings -- 4.2 Evaluation Metrics ; -- 5 Results and Discussions -- 5.1 Main Results -- 5.2 Ablation Study -- 5.3 Speed Study -- 6 Conclusion -- References -- ACT-R Theory Can Promote Personality Analysis of Social Network Subjects -- 1 Motivation -- 2 Theoretical Foundation -- 2.1 MBTI Personality Model -- 2.2 The Personality Cognition Approach Based on Social Text Data -- 3 Cognitive Processes of MBTI Personality Based on the ACT-R -- 3.1 Generative Rule Design (GRD) -- 3.2 Module Design -- 3.3 Cognitive Processes of Personality -- 4 Experimental Tests -- 5 Future Expectations and Limitations -- 6 Summary -- References -- MSMD: A Multi-Stage Meta Distillation Strategy for Cross-Lingual Natural Language Understanding -- 1 Introduction -- 2 Data and Methods -- 2.1 Dataset -- 2.2 Methodology -- 3 Results -- 3.1 Experimental Setup -- 3.2 Baselines -- 3.3 Evaluation -- 3.4 Experimental Results -- 4 Discussion -- 5 Conclusions -- References -- A Bilingual Templates Data Augmentation Method for Low-Resource Neural Machine Translation -- 1 Introduction -- 2 Related Work -- 3 Approach -- 3.1 Template Generation Algorithm -- 3.2 Data Augmentation Strategy -- 4 Experiment and Settings -- 4.1 Datasets -- 4.2 Pre-processing and Settings -- 5 Results -- 5.1 Performance on Different Translation Tasks -- 5.2 Comparison with Other Data Augmentation (DA) Methods -- 5.3 Performance by Sentence Length -- 5.4 Ablation Study -- 5.5 Comparison of BLEU Score and Training Loss -- 6 Conclusion -- References.; Simple Techniques for Enhancing Sentence Embeddings in Generative Language Models -- 1 Introduction -- 2 Methodology -- 2.1 EOL in the Context of Discriminative Models -- 2.2 EOL in the Context of Generative Models -- 2.3 Pretended CoT and Knowledge Enhancement -- 3 Experiments -- 3.1 Efficacy on 7B-Scale Generative PLMs -- 3.2 Computational Cost Comparison -- 3.3 PLM Performance Across Different Scales -- 4 Analysis -- 4.1 Alignment and Uniformity -- 4.2 Attention Allocation -- 5 Conclusion -- References -- DocBAN: An Efficient Biaffine Attention Network for Document-Level Named Entity Recognition -- 1 Introduction -- 2 Related Work -- 3 Proposed Method -- 3.1 Task Formalization -- 3.2 Encoder Module -- 3.3 Span Representation Module -- 3.4 Classification Module -- 4 Experiments -- 4.1 Experimental Setup -- 4.2 Main Results and Analysis -- 5 Conclusion -- References -- A Knowledge Graph Question Answering Approach Based on Graph Attention Networks and Relational Path Encoding -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Encoding Module -- 3.2 Inference Module -- 4 Experimentation -- 4.1 Results and Analysis -- 4.2 Ablation Study -- 5 Conclusion -- References -- News Sequence Recommendation Model with Dual-View Category Enhancement -- 1 Introduction -- 2 Related Work -- 3 Recommendation Method -- 3.1 Model Framework -- 3.2 Generation of News Representation Vectors -- 3.3 Generation of User Representation Vector -- 4 Experiments -- 4.1 Dataset and Metrics -- 4.2 Comparative Models -- 4.3 Results and Analysis -- 4.4 Ablation Study -- 4.5 Hyperparameter Analysis -- 5 Conclusion -- References -- CRFLOE: Context Region Filter and Relation Word Aware for Document-Level Relation Extraction -- 1 Introduction -- 2 Related Work -- 2.1 Relation Extract -- 2.2 Document-Level Relation Extraction -- 3 Proposed Method -- 3.1 Problem Formulation.; 3.2 Representation Learning -- 3.3 Context Region Aware -- 3.4 Relation Classifier -- 3.5 Relation Word Aware -- 4 Experiments -- 4.1 Dataset Statistics -- 4.2 Experiment Setup -- 4.3 Results on DocRE -- 4.4 Results on Re-DocRE -- 4.5 Results on CDR and GDA -- 4.6 Performance Analysis -- 5 Conclusion -- References -- Multi-task Learning for Hyper-Relational Knowledge Graph Completion -- 1 ; Introduction -- 2 Related Work -- 2.1 Hyper-relational Knowledge Graph Completion -- 2.2 Dynamic Loss Balancing -- 3 Preliminaries -- 3.1 Definition of Hyper-Relational Knowledge Graph -- 3.2 Definition of Hyper-Relational Knowledge Graph Completion -- 3.3 Definition of Task Difficulty -- 4 Method -- 4.1 Triple Encoder and Qualifier Encoder -- 4.2 Coarse-Grained Transformer Encoder -- 4.3 Fined-Grained Transformer Encoder -- 4.4 Training of MTL-HKGC Based on Task Difficulty -- 5 Experiment -- 5.1 Basic Setting -- 5.2 Entity Prediction Results -- 5.3 Relation Prediction Results -- 5.4 Effects of Multi-task Learning Architecture and Dynamic Loss Weighting -- 6 Conclusion -- References -- Instruct-Code-Llama: Improving Capabilities of Language Model in Competition Level Code Generation by Online Judge Feedback -- 1 Introduction -- 2 Dataset -- 2.1 Dataset Source -- 2.2 Rating Strategy -- 2.3 Data Split -- 3 Methods and Training Details -- 3.1 High-Level Methodology -- 3.2 Training -- 4 Results -- 4.1 Results on APPS -- 4.2 Loop Keywords -- 5 Conclusions -- References -- ICLFP-NMT: Neural Machine Translation for ICL Flexible Prompt -- 1 Introduction -- 2 Relate Works -- 3 Approach -- 3.1 Instances Generation Methods Based on Soft Prototypes -- 3.2 Virtual Template Automated Construction Methods -- 4 Experiments -- 4.1 Dataset and Experimental Methods -- 4.2 Baseline and Comparison Models -- 4.3 Experimental Results and Analysis -- 4.4 Ablation Experiment -- 5 Conclusion.; References -- Optimizing Multi-feature Dependent Story Generation Model -- 1 Introduction -- 2 Related Work -- 2.1 Text Genertion -- 2.2 Common Sense Knowledge -- 3 Method -- 3.1 Knowledge Retrieval -- 3.2 Dual-Encoder -- 3.3 Knowledge Management Decoder -- 4 Experiments -- 4.1 Datasets and Common Sense Knowledge -- 4.2 Experimental Details -- 4.3 Baselines and Evaluation Methods -- 4.4 Automatic Evaluation -- 4.5 Manual Evaluation -- 4.6 Comparative Experiments of Knowledge -- 4.7 Case Study -- 5 Conclusions -- References -- Improving Empathetic Response Generation by Emotion Recognition and Information Filtration -- 1 Introduction -- 2 Related Works -- 3 Proposed Method -- 3.1 Task Definition -- 3.2 Emotion Recognition -- 3.3 Information Filtration -- 3.4 Response Generation -- 3.5 Loss Function -- 4 Experiments -- 4.1 Dateset and Baselines -- 4.2 Evaluation Metrics -- 4.3 Results -- 4.4 Ablation Study -- 4.5 Case Study -- 4.6 Conclusion -- References -- Implementing Dialogue Emotion Recognition via Recursive Propagation of Explicit Emotions -- 1 Introduction -- 2 Related Work -- 3 Methodology -- 3.1 Problem Definition -- 3.2 Model Overview -- 3.3 Feature Extraction Module -- 3.4 Emotional Contagion Module -- 3.5 Emotional Inheritance Module -- 3.6 Classification and Training -- 4 Experiments -- 4.1 Datasets and Evaluation Indicators -- 4.2 Baselines -- 4.3 Implementation Details -- 4.4 Results and Analysis -- 4.5 Ablation Study -- 4.6 Visualization of Dialogue Emotion Mapping -- 5 Conclusion -- References -- Refining ChatGPT for Document-Level Relation Extraction: A Multi-dimensional Prompting Approach -- 1 Introduction -- 2 Method -- 2.1 Prompt Template -- 2.2 Multi Dimensional Prompting -- 3 Experiment and Analysis -- 3.1 Experimental Setup -- 3.2 Baseline -- 3.3 Main Results -- 3.4 Ablation Study -- 3.5 Hyperparameter Analysis.; 3.6 Analysis of Bottlenecks and Potential Improvements -- 3.7 Conclusion -- References -- AugSBertChat: User Feedback-Enhanced QA with Sentence-RoBERTa -- 1 Introduction -- 2 Background and Preliminaries -- 2.1 Problem Definition -- 2.2 Evaluation Metric -- 3 AugSBertChat -- 3.1 Scoring Model -- 3.2 Sentence-RoBERTa Enhanced by In-Domain Symmetric Semantic Search -- 3.3 Optimizing ; Response Generation Using Scoring Model -- 4 Experiments -- 4.1 Experiment Settings -- 4.2 Experimental Results and Analysis -- 5 Conclusion -- References -- Overlapping Entity Relation Extraction Based on Syntactic Dependency Tree and Multi-dimensional Corner Marking Strategy -- 1 Introduction -- 2 Related Work -- 3 Method -- 3.1 Semantic Feature Extraction -- 3.2 Syntactic Feature Extraction -- 3.3 Feature Fusion -- 3.4 Multi-dimensional Corner Marking Strategy -- 3.5 Score-Based Classifier -- 4 Experiments -- 4.1 Datasets and Evaluation Indicators -- 4.2 Experimental Results and Analysis -- 4.3 Ablation Experiment -- 4.4 Entity Relation Overlap Problem -- 5 Conclusion -- References -- Document-Level Relation Extraction with Additional Evidence and Entity Type Information -- 1 Introduction -- 2 Related Work -- 3 Model -- 3.1 Problem Formulation -- 3.2 Encoder -- 3.3 STA Layer -- 3.4 Context Pooling and Evidence Fusion -- 3.5 Entity Type Constraint -- 3.6 Classifier and Loss Function -- 3.7 Inference Stage Fusion -- 4 Experiment -- 4.1 Dataset -- 4.2 Evaluation Metrics -- 4.3 Implementation Details -- 4.4 Experiment Result -- 4.5 Ablation Study -- 4.6 The Performance of Entity Type Constraint -- 5 Conclusion -- References -- Multi-level Contrastive Learning for Keyphrase Generation -- 1 Introduction -- 2 Related Work -- 2.1 Keyphrase Generation -- 2.2 Contrastive Learning -- 3 Methodology -- 3.1 Problem Definition and Framework -- 3.2 Backobone.; 3.3 Phrase-Level Contrastive Learning.
• Sommario/riassunto: This 6-volume set LNAI 14875-14880 constitutes - in conjunction with the 13-volume set LNCS 14862-14874 and the 2-volume set LNBI 14881-14882 - the refereed proceedings of the 20th International Conference on Intelligent Computing, ICIC 2024, held in Tianjin, China, during August 5-8, 2024. The total of 863 regular papers were carefully reviewed and selected from 2189 submissions. The intelligent computing annual conference primarily aims to promote research, development and application of advanced intelligent computing techniques by providing a vibrant and effective forum across a variety of disciplines. This conference has a further aim of increasing the awareness of industry of advanced intelligent computing techniques and the economic benefits that can be gained by implementing them. The intelligent computing technology includes a range of techniques such as Artificial Intelligence, Pattern Recognition, Evolutionary Computing, Informatics Theories and Applications, Computational Neuroscience & Bioscience, Soft Computing, Human Computer Interface Issues, etc.