Artificial intelligence and mobile services - AIMS 2022 : 11th international conference, held as part of the services conference federation, SCF 2022, Honolulu, HI, USA, December 10-14, 2022 : proceedings

This book constitutes the proceedings of the 11th International Conference on Artificial Intelligence and Mobile Services, AIMS 2022, held as Part of the Services Conference Federation, SCF 2022, held in Honolulu, HI, USA, in December 2022. The 10 full papers presented in this volume were carefully reviewed and selected from 22 submissions. The International Conference on AI & Mobile Services (AIMS 2022) aims at providing an international forum that is dedicated to exploring different aspects of AI (from technologies to approaches and algorithms) and mobile services (from business management to computing systems, algorithms, and applications) to promoting technological innovations in research and development of mobile services, including, but not limited to, wireless & sensor networks, mobile & wearable computing, mobile enterprise & eCommerce, ubiquitous collaborative & social services, machine-to-machine & Internet-of-things clouds, cyber-physical integration, and big data analytics for mobility-enabled services. Preface Organization Services Society Services Conference Federation (SCF) Contents Research Track Push-Based Forwarding Scheme Using Fuzzy Logic to Mitigate the Broadcasting Storm Effect in VNDN 1 Introduction 2 Vehicular Named Data Networking (VNDN) 2.1 Content Store (CS) 2.2 Pending Interest Table (PIT) 2.3 Forwarding Information Base (FIB) 2.4 Consumer Vehicle 2.5 Producer Vehicle 3 Push-Based Data Forwarding in VNDN 4 Proposed Push-Based Data Forwarding Scheme with Fuzzy Logic 4.1 K-Means Clustering 4.2 Selection of Cluster Head (CH) Using Fuzzy Logic 4.3 Proposed Data Packet Format 4.4 Proposed Scheme for Producer 4.5 Proposed Scheme for Consumer 4.6 Critical Data Forwarding Procedure by the Proposed Scheme 5 Simulation Environment and Results 6 Conclusion References DCRNNX: Dual-Channel Recurrent Neural Network with Xgboost for Emotion Identification Using Nonspeech Vocalizations 1 Introduction 2 Related Works 3 Proposed Method 3.1 Dual-Channel Neural Network Model 3.2 Introducing Attention Mechanism in Two-Channel Model 3.3 XGBoost Classifier 3.4 Model Fusion Using L2 Norm 4 Experiments 4.1 Datasets Used 4.2 Experimental Setup 4.3 Experimental Results 4.4 Introduce Attention Mechanism 4.5 Data Augmentation 5 Conclusion References STaR: Knowledge Graph Embedding by Scaling, Translation and Rotation 1 Introduction 2 Related Work 3 Methodology 3.1 Background Knowledge 3.2 The Proposed STaR Model 3.3 Discussions 4 Experiments 4.1 Experiments Settings 4.2 Main Results 5 Analysis 5.1 Further Comparison with ComplEx 5.2 Imbalance Ratio Among KGs 5.3 Improvements on WN18RR Come from Modeling Non-commutativity Pattern 6 Conclusion References Application Track Frequently Asked Question Pair Generation for Rule and Regulation Document 1 Introduction 2 Related Work 2.1 QA Pair Generation 2.2 FAQ Pair Generation 3 Data Collection and Analysis 3.1 Data Collection 3.2 Data Analysis 4 Methodology 4.1 Rule-Based FAQ Pair Generation 4.2 Pipeline Framework 5 Experiment 5.1 Experiment Setting 5.2 Evaluation Metrics 5.3 Analysis Experiment 5.4 Human Evaluation 5.5 Case Study 6 Conclusion References Chinese Text Classification Using BERT and Flat-Lattice Transformer 1 Introduction 2 Related Work 2.1 Traditional and Embedding-Based Text Classification 2.2 Neural Network Text Classification 2.3 Chinese Text Classification 2.4 Transformer Related Theory 3 Approaches 3.1 Converting Lattice into Flat Structure 3.2 Relative Position Encoding of Spans 3.3 Classifier and Optimization 4 Experiments 4.1 Experimental Setup 4.2 Overall Performance 5 Conclusion References Indicator-Specific Recurrent Neural Networks with Co-teaching for Stock Trend Prediction 1 Introduction 2 Related Work 2.1 Technical Indicators 2.2 Stock Trend Prediction Model 2.3 Co-teaching for Noisy Data 3 Dataset 3.1 Data Acquisition 3.2 Data Processing 3.3 Dataset Construction 4 Methodology 4.1 Task Definition 4.2 Model Overview 4.3 Feature Extraction Module 4.4 Feature-Based Attention Module 4.5 Prediction Module 4.6 Co-teaching 5 Experiment 5.1 Experimental Setting 5.2 Results and Analysis 5.3 Ablation Study 5.4 Case Study 6 Conclusion References SATMeas - Object Detection and Measurement: Canny Edge Detection Algorithm 1 Introduction 2 Methods 2.1 General Idea 2.2 Platform and Technology Used 2.3 Object Detection and Object Measurement 3 Results 3.1 Advantages 3.2 Disadvantages 4 Conclusion References Multi-Classification of Electric Power Metadata based on Prompt-tuning 1 Introduction 2 Related Works 2.1 Short Text Classification 2.2 Prompt 2.3 Verbalizer 3 Methodology 3.1 Task Formulation 3.2 Pre-trained Model with Electric Power Knowledge 3.3 Deep Prompt-Tuning Module 3.4 Exterior Knowledge-Based Verbalizer 4 Experiment 4.1 Dataset 4.2 Parameter Settings 4.3 Evaluation Metrics 4.4 Experiments 5 Conclusion References Dual-Branch Network Fused with Attention Mechanism for Clothes-Changing Person Re-identification 1 Introduction 2 Related Work 3 Method 3.1 Network 3.2 Loss Function 4 Experiments 4.1 Dataset 4.2 Experimental Details 4.3 Experimental Results 4.4 Visual Analysis 5 Conclusion References Infant Cry Classification Based-On Feature Fusion and Mel-Spectrogram Decomposition with CNNs 1 Introduction 2 Spectrograms vs. Mel-Spectrograms 3 Model Architecture 3.1 Spectrogram Feature Extractor 3.2 Mel-Spectrogram Feature Extractor 4 Experimental Setup and Results 4.1 Datasets 4.2 Experimental Setup 4.3 Results and Analysis 5 Conclusion References Author Index