Computer Supported Cooperative Work and Social Computing: 17th CCF Conference, ChineseCSCW 2022, Taiyuan, China, November 25–27, 2022, Revised Selected Papers, Part I

Description

This two-volume set constitutes the refereed proceedings of the 17th CCF Conference on Computer Supported Cooperative Work and Social Computing, ChineseCSCW 2022 held in Taiyuan, China, during November 25–27, 2022. The 60 full papers and 30 short papers included in this two-volume set were carefully reviewed and selected from 211 submissions. They were organized in topical sections as follows: answer set programming; Social Media and Online Communities, Collaborative Mechanisms, Models, Approaches, Algorithms and Systems; Crowd Intelligence and Crowd Cooperative Computing; Cooperative Evolutionary Computation and Human-like Intelligent Collaboration; Domain-Specific Collaborative Applications. Preface Organization Contents – Part I Contents – Part II Social Media and Online Communities Multi-step Ahead PM2.5 Prediction Based on Hybrid Machine Learning Techniques 1 Introduction 2 Feature Selection 2.1 Candidate Feature Pool Construction Based on PSR Method and NWP Data 2.2 Optimal Feature Selection Based on RReliefF Algorithm 3 The Proposed Hybrid Prediction Model 3.1 Subset Division Based on K-means Clustering Algorithm 3.2 Optimization of ANFIS Network Model Based on PSO Algorithm 4 Analysis of Results 4.1 Feature Selection Analysis 4.2 Feature Selection Analysis 5 Conclusion References A Joint Framework for Knowledge Extraction from Flight Training Comments 1 Introduction 2 Related Work 2.1 Keyword Extraction 2.2 Aspect Based Sentiment Analysis 3 Model Architecture 3.1 Task Definition 3.2 Grid Tagging Scheme 3.3 Encoding 3.4 Inference Strategy 3.5 Decoding 4 Experiment 4.1 Dataset 4.2 Experimental Setup 4.3 Experimental Results 5 Conclusion References ScholarRec: A User Recommendation System for Academic Social Network 1 Introduction 2 Related Work 2.1 Collaborative Filtering Based Recommendation System 2.2 Content Based Recommendation System 2.3 Implicit Semantics Based Recommendation System 3 System Design 3.1 System Architecture 3.2 Key Modules 3.3 Key Processes 3.4 Critical Problem Solving 4 Implementation and Evaluation 4.1 SCHOLAT Social Network Dataset 4.2 Model Evaluation 4.3 Performance Comparison 4.4 Demonstration Effect 5 Conclusion References Incremental Evolutionary Community Discovery Method Based on Neighbor Subgraph 1 Introduction 2 Related Work 3 Incremental Evolutionary Community Discovery Method Based on Neighbor Subgraph 3.1 Core Node Discovery Based on Neighbor Subgraph 3.2 Community Evolution Based on Dynamic Contribution 4 Experiment 4.1 Static Real Network 4.2 Dynamic Real Network Results 5 Conclusion References Video Rumor Classification Based on Multi-modal Theme and Keyframe Fusion 1 Introduction 2 Related Work 2.1 Multi-modal Information Fusion 2.2 Contrast Learning Between Models 2.3 Enhance Multi-modal Information 3 Method 3.1 Modal Feature Extraction 3.2 Theme Features and Key-Frame Feature 3.3 Modal Adjustment Mechanism 4 Experiment 4.1 DataSet 4.2 Implementation Details 4.3 Modal Combination Analysis 4.4 Model Comparison Experiment 4.5 Ablation Experiment 5 Conclusion References Association Rule Guided Web API Complementary Function Recommendation for Mashup Creation: An Explainable Perspective 1 Introduction 2 Related Work 3 Motivation 3.1 Necessity 3.2 Formulation 4 Methodology 4.1 Web API Complementary Function Recommendation System 4.2 Mining Association Rules 4.3 Association Rule Analyses 4.4 Experiments 5 Application Scenarios 5.1 Typical Application Scenarios 6 Conclusion References Globally Consistent Vertical Federated Graph Autoencoder for Privacy-Preserving Community Detection 1 Introduction 2 Related Work 2.1 Privacy-preserving Community Detection 2.2 Federated Learning on Graphs 3 Preliminaries 3.1 Problem Definition 3.2 Standard Graph Autoencoder 4 Globally Consistent Vertical Federated Graph Autoencoder 4.1 Design of GCVFGAE 5 Experiments 5.1 Datasets 5.2 Baseline Algorithms 5.3 Experimental Settings and Evaluation Metrics 5.4 Consistency Experiment 5.5 Ablation Experiment 5.6 Accuracy Experiment 5.7 Running Time Experiment 6 Conclusions References Research on User Personality Characteristics Mining Based on Social Media 1 Introduction 2 Related Research 3 ATCNN-BiGRU Model 3.1 Text Processing 3.2 Feature Noise Reduction and Fusion 3.3 Word Vector Attention Mechanism 3.4 Convolutional Neural Network Module Combined with Attention Mechanism 3.5 Overall Model Architecture 4 Experiments 4.1 Analysis of Experimental Results 4.2 Loss Function Analysis 4.3 Analysis of the Effectiveness of Attention Mechanism 5 Summary References A Unified Stream and Batch Graph Computing Model for Community Detection 1 Introduction 2 Related Work 2.1 Batch Large-Scale Graph Computation 2.2 Stream Large-Scale Graph Computation 2.3 Unified Stream and Batch Processing 3 Unified Stream and Batch Graph Computing Model 3.1 The Framework of USBGM 3.2 Unified Stream and Batch Processing and Graph Model Cache Update 3.3 Time Decay Strategy 4 Application of the USBGM 4.1 Application of Graph Operators Based on USBGM 4.2 Application of Community Detection Algorithm Based on USBGM 5 Experiments 5.1 Datasets Description 5.2 Baseline Algorithms 5.3 Evaluation Metrics 5.4 Accuracy Experiment 5.5 Time Decaying Effect Experiment 6 Conclusion References A Feature Fusion-Based Service Classification Approach for Collaborative Development 1 Introduction 2 The SCAFF Approach 2.1 Service Invocation Graph Generation 2.2 Service Feature Extraction 2.3 Feature Fusion 2.4 Service Classification 3 Experiments 3.1 Experiment Setup 3.2 Dataset 3.3 Evaluation Metrics 3.4 The Comparative Experiment 3.5 The Ablation Experiment 3.6 Results 4 Conclusion and Future Work References Requirements Classification and Identification Approach for E-Collaboration Systems 1 Introduction 2 Method 2.1 Research Framework 2.2 Background 2.3 Definition of Requirement Categories for E-Collaboration Systems 2.4 Dataset Pre-processing 2.5 Requirement Analysis and Feedback 3 Experiments and Effect Analysis 3.1 Comparison Method 3.2 Experiment Setup 3.3 Experimental Effects and Analysis 4 Conclusion and Future Work References Community Evolution Tracking Based on Core Node Extension and Edge Variation Discerning 1 Introduction 2 Related Work 2.1 Instant-Optimal Methods 2.2 Temporal Trade-Off Methods 2.3 Cross-Time Methods 3 The Proposed Algorithm 3.1 Preliminaries 3.2 Framework of CETCE 3.3 Initial Community Detection 3.4 Incremental Community Updates and Evolution Tracking 3.5 Time Complexity Analysis 4 Experiments 4.1 Datasets 4.2 Evaluation Metrics 4.3 Baseline Algorithms 4.4 Accuracy Experiments 5 Conclusion References University Knowledge Graph Construction Based on Academic Social Network 1 Introduction 2 Related Work 3 Construction of University Knowledge Graph 3.1 Ontology Construction 3.2 Knowledge Extraction 3.3 Knowledge Graph Storage 4 Application Analysis of the University Knowledge Graph 5 Conclusion References Country-Level Collaboration Patterns of Social Computing Scholars 1 Introduction 2 Related Work 2.1 Country-Level Co-authorship Collaboration Network 2.2 Country-Level Collaboration Preference 3 Data and Methods 3.1 Data Collection and Collaboration Network Construction 3.2 Methods 4 Country Collaboration Preference 4.1 Different Types of Countries in The Collaboration Network 4.2 Attract Index and Group-wise Attract Index 4.3 Collaboration Preferences of Different Types of Countries 5 Academic Influence of the Collaboration Preference 5.1 Hypothesis 5.2 Model Results 6 Conclusions and Future Work References An Intelligent Mobile System for Monitoring Relapse of Depression 1 Introduction 2 Literature Review 2.1 Acoustic Detection 2.2 Semantic Analysis 3 Data Collection and Preprocessing 3.1 Data Collection 3.2 Acoustic Data Preprocessing 3.3 Semantic Text Preprocessing 3.4 Semantic Dictionary Construction 4 Machine Learning 4.1 Learning Model 4.2 Learning Results 5 System Design 5.1 System Architecture 5.2 System Operation 6 Summary and Discussion References Fine-Grained Sentiment Analysis of Online-Offline Danmaku Based on CNN and Attention 1 Introduction 2 Related Work 2.1 Short Text Classification 2.2 Sentiment Analysis 3 Proposed Method 3.1 Classification Criteria of Danmaku 3.2 Sentiment Classifier 3.3 Intent Classifier 3.4 Visualization of Sentiment Intensity 4 Experiment 4.1 Dataset 4.2 Model Comparison 4.3 The Generation of Sentiment Curves 5 Sentiment Analysis 5.1 Online Danmaku 5.2 Offline Danmaku 6 Conclusion References Ramp Merging of Connected Vehicle With Virtual Platooning Control 1 Introduce 2 Ramp Merging Algorithm Based on the Virtual Formation 2.1 Combined Zoning and Model Assumptions 2.2 Optimization of Traffic Order in the Pre-merge Area 2.3 Merging Method of Virtual Formation in the Inbound Area 3 Simulation and Analysis of Ramp Confluence 3.1 Ramp Confluence Simulation Scenario Construction 4 Simulation Results and Analysis of All CAV Scenarios References Community Detection Based on Enhancing Graph Autoencoder with Node Structural Role 1 Introduction 2 Related Work 3 CDESR Algorithm 3.1 Preliminaries 3.2 Node Structure Role Recognition 3.3 Topology Information Extraction 3.4 Enhancement Matrix Construction 3.5 Community Detection 4 Experiments 4.1 Datasets 4.2 Baseline Algorithms 4.3 Evaluation Metrics 4.4 Parameter Settings 4.5 Parameter Experiment 4.6 Accuracy Experiment 4.7 Ablation Study 5 Conclusion References Representation of Chinese-Vietnamese Bilingual News Topics Based on Heterogeneous Graph 1 Introduction 2 Methodology 2.1 Encoder 2.2 Decoder 3 Experimental Results and Analysis 3.1 Evaluation Metric 3.2 Datasets 3.3 Baseline Model 3.4 Comparative Experimental Analysis 3.5 Ablation Experiment 3.6 Case Analysis 4 Conclusions and Future Work References Convolutional Self-attention Network for Sequential Recommendation 1 Introduction 2 Related Work 2.1 General Recommendation 2.2 Sequential Recommendation 3 Model Design and Architecture 3.1 Problem Formulation 3.2 Model Architecture 3.3 Convolutional Layer 3.4 Transformer Layer 3.5 Fine-Tuning Technique Based on Triplet Loss 4 Experiments and Result 4.1 Dataset Description 4.2 Dataset Preparation 4.3 Evaluation Metrics 4.4 Comparative Methods and Implementation Details 4.5 Experimental Setup 4.6 Overall Performance Comparison 4.7 Impact of Margin of Triplet Loss 4.8 Ablation Study 5 Conclusion References Towards Using Local Process Mining to Analyse Learning Behavior Pattern 1 Introduction 2 Preliminaries 3 Analyse Learning Behavior Pattern by Local Process Mining 3.1 LBPA Method 3.2 Application Analysis 4 Conclusion References Collaborative Mechanisms, Models, Approaches, Algorithms and Systems Memory-Effective Parallel Mining of Incremental Frequent Itemsets Based on Multi-scale 1 Introduction 2 Related Work 2.1 Parallel Incremental FIM 2.2 Spark Caching Mechanism 3 Incremental Frequent Itemsets Mining Based on Multi-scale 3.1 Related Definitions of Multi-scale 3.2 The Process of Incremental Frequent Itemsets Mining Based on Multi-scale 4 Multi-scale Incremental Frequent Itemsets Mining Based on Spark 4.1 Introduction to Parallel Algorithms 4.2 The RDD Conversion Process of Parallel FPMSIM Algorithm 5 Low Cost Double Execution Replacement Policy (LCDERP) 6 Experimental Results and Analysis 6.1 Experimental Setup 6.2 Experimental Results and Analysis 7 Conclusion and Future Work References An AST-Based Collaborative Discussion Tool for the MOOC Environment 1 Introduction 2 Related Work 3 Simulated Collaboration Modes 3.1 Reading Supplementary Mode 3.2 Cumulative Discussion Mode 3.3 Guidance Mode 3.4 Replay Mode 4 An AST-Based Approach for the Replay Mode 4.1 Data Model 4.2 Consistency Model 4.3 Operation Model 5 Implementation and Evaluation 6 Conclusion and Future Work References DQN-Based Comprehensive Consumption Minimization on Calculation Offloading in Mobile Edge Computing 1 Introduction 2 System Model for MSDE 2.1 Local Computing 2.2 Computing Offload 2.3 Latency and Energy Consumption 3 Problem Formulation 4 DQN-Based Offloading Strategy 4.1 MDP Process 4.2 DQN Algorithm 5 Experimental Simulation 6 Summary and Outlook References Stochastic Task Offloading Problems for Edge Computing 1 Introduction 2 Related Work 3 Problem Description 4 Proposed Methods 4.1 Simulated Annealing Based Online Offloading Framework 4.2 Simulated Annealing Based Algorithm 4.3 ARR 4.4 OFFLOADING 5 Experimental Evaluation 6 Conclusion References Container-Driven Scheduling Strategy for Scientific Workflows in Multi-vCPU Environments 1 Introduction 2 Model and Problem Definition 2.1 Workflow Model 2.2 Resource Model 2.3 Problem Definition 3 Container-Based Segmented Workflow Scheduling Strategy 3.1 Generation of Task Scheduling Sequence Table 3.2 Segmented Scheduling Approach 3.3 Modified Heuristic Algorithm ADPSOGA 3.4 Container Scaling Scheme 4 Experiments and Analysis 4.1 Experimental Setup 4.2 Segmented Scheduling Effectiveness Verification 4.3 Comparison of Algorithms 5 Conclusion References A Segmented Path Heuristic Recovery Algorithm for WSNs Based on Mobile Sink 1 Introduction 2 Related Work 3 Network Model and Proposed Algorithm 3.1 Network Model 3.2 Dynamic Segmented Path Heuristic Recovery 4 Experimental Simulation and Result Analysis 4.1 Feasibility of Segmentation Path Recovery Method 4.2 Energy Consumption Analysis of Segmented Path Algorithm 4.3 Delay Analysis of Path Recovery Algorithm 4.4 Effectiveness Analysis of WOA Reconstruction Algorithm 5 Conclusion References TRindex: Distributed Double-Layer Road Network Trajectory Index 1 Introduction 2 Related Work 3 TRindex Model 3.1 Linear Order Partition 3.2 TRindex Structure 3.3 Data Cache 3.4 Load Balancing 4 TRindex Index Operation 4.1 The Query of TRindex 4.2 The Update of TRindex 5 Performance Evaluation 5.1 TRindex Construction 5.2 TRindex Query 5.3 TRindex Insert 6 Conclusion References Sleep Scheduling for Enhancing the Lifetime of Three-Dimensional Heterogeneous Wireless Sensor Networks 1 Introduction 2 Related Work 3 System Model and Problem Definition 3.1 System Model 3.2 Problem Definition 4 Redundant Sensor Estimation Method 5 A Distributed Sleep Scheduling Algorithm 6 Simulation Results 6.1 Comparison Results 6.2 Impact of Border Effects and Probabilistic Sensing Model 7 Conclusion References CoSBERT: A Cosine-Based Siamese BERT-Networks Using for Semantic Textual Similarity 1 Introduction 2 Related Work 2.1 Semantic Textual Similarity 2.2 Related Models 3 Model 3.1 CoSBERT 3.2 Loss Function 3.3 Training Details 4 Experiment 4.1 Datasets and Metrics 4.2 Pre-trained Models 4.3 Evaluation 4.4 Ablation Study 5 Conclusion References Towards Heterogeneous Federated Learning 1 Introduction 2 Preliminaries 2.1 Definition 2.2 Analysis 3 The Proposed Heterogeneous FL Taxonomy 3.1 Federal Learning Strategies with Heterogeneous Device 3.2 Federal Learning Strategies with Heterogeneous Network 3.3 Federal Learning Strategies with Heterogeneous Data 3.4 Federal Learning Strategies with Heterogeneous Model 4 Heterogeneous Federal Learning Evaluation Methods 4.1 Theoretical Analysis of HFL 4.2 Experimental Evaluation Metrics 5 Promising Future Research Directions References A Graph-Based Efficient Service Composition Method for Computer Aided Engineering (CAE) 1 Introduction 2 N-WSPR Algorithm 2.1 Service Discovery 2.2 Service Selection 2.3 Service Ordering 2.4 Analysis of Algorithms 3 Experiments 3.1 Operating Environment 3.2 Comparison of the Running Time of Different Algorithms 3.3 The Influence of the Number of Services on the Running Time of the Algorithm 3.4 Analysis of the Results 4 Conclusion References Privacy-Preserving Federated Learning Framework in Knowledge Concept Recommendation 1 Introduction 2 Related Work 3 Methodology 3.1 Problem Formulation 3.2 Recommendation in Client 3.3 Federated Setting in Server 4 Experiment 4.1 Dataset and Experiment Setting 4.2 Detailed Analysis 5 Conclusion References RCPM: A Rule-Based Configurable Process Mining Method 1 Introduction 2 Related Work 3 RCP Model: A Rule-based Configurable Process Model 4 RCPM: RCP Model Mining Algorithm 4.1 Process Tree 4.2 Steps 4.3 Computing Branch Frequency 4.4 Cutting Branches with Low Frequency 5 Experiment 5.1 Experiment Setup 5.2 Method Comparison 5.3 Impact of Threshold Values 6 Conclusion and Future Work References Popularity Bias Analysis of Recommendation Algorithm Based on ABM Simulation 1 Introduction 1.1 Methods Based on Static Datasets 1.2 Method Based on Multi-round Interaction 2 Related Work 2.1 Recommendation Algorithm Popularity Bias 3 Construction of ABM-Based Recommendation System Simulation Framework 3.1 System Build 3.2 System Operation Details 4 Experimental Validation and Data Analysis 4.1 Experimental Tools and Environment 4.2 Metrics and Parameter Setting 4.3 Popularity Bias Analysis of Two Algorithms 5 Summary and Outlook References Cloud-Edge Collaborative Task Scheduling Mechanism Based on Improved Parameter Adaptation Particle Swarm Optimization Algorithm 1 Introduction 2 The Problem’s Description and the Modeling 2.1 Problem Description 2.2 Model Building 3 Task Scheduling Based on Parameter Adaptive Particle Swarm Optimization Algorithm 4 Experiments and Evaluation 4.1 Evaluation of Algorithm Performance 4.2 Task Scheduling Simulation Experiment 5 Conclusion References An Approach to Assessing the Health of Opensource Software Ecosystems 1 Introduction 2 Related Work 2.1 Software Ecosystem Health 2.2 Open-Source Software Ecosystem Health 3 Qualitative and Quantitative Evaluation Methods of Open-Source Software Ecosystem Health 3.1 Data Acquisition Method 3.2 Qualitative Evaluation Method 3.3 Quantitative Evaluation Model 3.4 Score Conversion 4 System Design and Implementation 4.1 System Design 4.2 Back-End Implementation 4.3 Web-Based Front-End Implementation 5 Experiment on Open-Source Software Ecosystem Health Evaluation 5.1 Data Acquisition and Score Calculation 5.2 Result Analysis 6 Conclusion References Topic Discovery in Scientific Literature 1 Introduction 2 Related Work 3 Research Hot Word Extraction Based on Node Centrality 3.1 The Construction of Word Network 3.2 Node Centrality Calculation and Hot Words Selection 4 Structural Topic Discovery and Consistence Analysis 4.1 Structural Topic Discovery 4.2 Consistence Analysis 5 Experimental Results and Analysis 6 Conclusion References Multi-agent Adversarial Reinforcement Learning Algorithm Based on Reward Query Attention Mechanism 1 Introduction 2 Background 2.1 Deep Q-Learning 2.2 Value Decomposition Network 2.3 Batch Regularization 2.4 Attention Mechanism 3 QMIX-NA Algorithm 3.1 Individual Agent Network 3.2 Attention Mechanism Network Layer 3.3 Mixture of Action Value Functions 4 Experiments 5 Conclusion References UAV Target Roundup Strategy Based on Wolf Pack Hunting Behavior 1 Introduction 2 Task Description 3 Hunting Algorithm Based on Wolf Pack Hunting Behavior 3.1 Brief Description of Hunting Behavior of Wolves 3.2 A Roundup Strategy Based on the Hunting Behavior of Wolves 3.3 Wolf Pack Hunting Behavior Combined with Reinforcement Learning 4 Simulation Experiment and Result Analysis 4.1 Experimental Simulation and Analysis of Wolves’ Hunting Behavior Roundup Strategy 4.2 Simulation Experiment of Wolf Pack Hunting Behavior Combined with Multi-agent Reinforcement Learning 5 Conclusion References Prediction of New Energy Vehicles via ARIMA-BP Hybrid Model 1 Introduction 2 Linear Prediction 2.1 Stationarity Test of Time Series 2.2 White Noise Series 2.3 Determine the Order of Model 3 Nonlinear Prediction 3.1 Determine the Number of Neurons and Hidden Layer Nodes 3.2 Select Activation Function 3.3 Data Normalization 3.4 Establish Prediction Model 4 Combined Forecast 4.1 ARIMA-BP Weighted Combination Model 4.2 ARIMA-BP Residual Optimization Combination Model 5 Experimental Analysis 5.1 Evaluation Indicators 5.2 Model Evaluation 6 Conclusion References Author Index