
Web and Big Data
Description
Alles über E-Books | Antworten auf Fragen rund um E-Books, Kopierschutz und Dateiformate finden Sie in unserem Info- & Hilfebereich.
The 4-volume set LNCS 14331, 14332, 14333, and 14334 constitutes the refereed proceedings of the 7th International Joint Conference, APWeb-WAIM 2023, which took place in Wuhan, China, in October 2023.
The total of 138 papers included in the proceedings were carefully reviewed and selected from 434 submissions. They focus on innovative ideas, original research findings, case study results, and experienced insights in the areas of the World Wide Web and big data, covering Web technologies, database systems, information management, software engineering, knowledge graph, recommend system and big data.
More details
Other editions
Additional editions

Content
- Intro
- Preface
- Organization
- Contents - Part III
- Adaptive Graph Attention Hashing for Unsupervised Cross-Modal Retrieval via Multimodal Transformers
- 1 Introduction
- 2 Related Work
- 3 Methodology
- 3.1 Notation and Problem Definition
- 3.2 Framework Overview
- 3.3 Objective Function
- 4 Experiments
- 4.1 Evaluation Metrics
- 4.2 Implementation Details
- 4.3 Comparison Results and Analysis
- 4.4 Ablation Study
- 4.5 Parameter Sensitivity Analysis
- 4.6 Convergence Testing
- 5 Conclusion
- References
- Answering Property Path Queries over Federated RDF Systems
- 1 Introduction
- 2 Related Work
- 3 Problem Formulation
- 4 The Proposed Method
- 4.1 Query Decomposition and Source Localization
- 4.2 Thompson-Based MinDFA Construction
- 4.3 Query Execution Strategy Base on B-DFS
- 5 Evaluation
- 5.1 Experimental Environment
- 5.2 Performance Comparison of Five Property Path Query Symbols
- 5.3 Performance and Resource Consumption of Different Matching Strategies of MinDFA
- 5.4 Performance Robustness of Five Property Path Query Symbols
- 6 Conclusion
- References
- Distributed Knowledge Graph Query Acceleration Algorithm
- 1 Introduction
- 2 Related Works
- 3 Offline Module for Distributed Construction of Indexes
- 3.1 MapReduce-Based Data Pre-processing
- 3.2 Coding-Oriented Construction of Distributed Hierarchical Clustering
- 4 Online Module for Distributed Parallel Processing of SPARQL Queries
- 4.1 Splitting and Loading of BitSet-Tree
- 4.2 Candidate Solution Acquisition
- 4.3 Shuffle
- 4.4 Merge Splicing of Candidate Vertices
- 5 Experiments
- 5.1 Experimental Settings
- 5.2 Results and Discussion
- 6 Conclusions
- References
- Truth Discovery of Source Dependency Perception in Dynamic Scenarios
- 1 Introduction
- 2 Problem Setting
- 2.1 Notation Definition
- 2.2 Task Description
- 3 Preliminary
- 3.1 Source Dependency Detection Based on Bayesian Model
- 3.2 Truth Discovery Framework Based on Optimization Model
- 4 Methodology
- 4.1 Source Dependency Detection in Dynamic Scenarios
- 4.2 Dynamic Incremental Model Framework
- 4.3 Truth Discovery with Source Dependency Perception
- 5 Experiment
- 5.1 Experimental Setup
- 5.2 The Results on Real-World Datasets
- 5.3 The Results on Synthetic Datasets
- 6 Related Work
- 7 Conclusion
- References
- Truth Discovery Against Disguised Attack Mechanism in Crowdsourcing
- 1 Introduction
- 2 Related Work
- 3 Preliminary
- 3.1 Disguised Attack Mechanism
- 3.2 Problem Formulation
- 3.3 Truth Discovery
- 4 Methodology
- 4.1 Behavior-Based Truth Discovery
- 4.2 Task Assignment Based on WAM
- 4.3 TD-DA Framework
- 5 Experiments
- 5.1 Experiment Setting
- 5.2 Verification of the Proportion of Malicious Workers
- 5.3 Experiment on Real-World Datasets
- 5.4 Experiment on Synthetic Datasets
- 6 Conclusion
- References
- Continuous Group Nearest Group Search over Streaming Data
- 1 Introduction
- 2 Preliminary
- 2.1 Related Works
- 2.2 Problem Definition
- 3 The Framework KMPT
- 3.1 The Basic Idea
- 3.2 The Initialization Algorithm
- 3.3 The Incremental Maintenance Algorithm
- 4 The Experiment
- 4.1 Experiment Settings
- 4.2 Performance Comparison
- 5 Conclusion
- References
- Approximate Continuous Skyline Queries over Memory Limitation-Based Streaming Data
- 1 Introduction
- 2 Preliminary
- 2.1 Related Works
- 2.2 Problem Definition
- 3 The Self-adaptive-based Framework -SEAK
- 3.1 The -CSS Definition
- 3.2 The Initialization Algorithm
- 3.3 The Incremental Maintenance Algorithm
- 3.4 The Partition-Based Optimization Algorithm
- 4 Performance Evaluation
- 4.1 Experiment Settings
- 4.2 Experimental Evaluation
- 5 Conclusion
- References
- Identifying Backdoor Attacks in Federated Learning via Anomaly Detection
- 1 Introduction
- 2 Related Work
- 2.1 Attacks on Model Faithfulness
- 2.2 Defenses Against Backdoor Attack
- 3 Preliminaries and Attack Formulation
- 3.1 Federated Learning
- 3.2 Threat Model
- 3.3 Choice of Backdoor Triggers
- 4 Methodology
- 4.1 Motivation
- 4.2 Segmenting Local Updates
- 4.3 Identifying Outliers in Fragments
- 4.4 Pruning Backdoored Participants
- 5 Experiments
- 5.1 Experimental Settings
- 5.2 Effectiveness of Our Defense
- 5.3 Comparison with Prior Arts
- 5.4 Effectiveness on Advanced Attacks
- 5.5 Ablation Study
- 6 Conclusion
- References
- PaTraS: A Path-Preserving Trajectory Simplification Method for Low-Loss Map Matching
- 1 Introduction
- 2 Related Work
- 2.1 Error-Bounded Line Simplification
- 2.2 Semantic-Preserving Trajectory Simplification
- 2.3 Analysis of Existing Work
- 3 Preliminaries
- 3.1 Basic Definitions
- 3.2 Methodology Analysis
- 4 Path-Preserving Trajectory Simplification
- 4.1 Overview of PaTraS
- 4.2 Preserving Shortest Paths
- 4.3 Candidates Pairing
- 4.4 Similarity Computation
- 4.5 Pairing Optimization
- 5 Experiments
- 5.1 Experimental Setup
- 5.2 Evaluation Oriented to Map-Matching
- 5.3 Parameter Sensitivity Study
- 6 Conclusion
- References
- Coordinate Descent for k-Means with Differential Privacy
- 1 Introduction
- 2 Related Work
- 3 Preliminary
- 3.1 k-Means
- 3.2 Coordinate Descent for k-Means
- 3.3 A Fast Version of CDKM
- 3.4 Differential Privacy
- 4 Proposed Our Method
- 4.1 Approximate CDKM
- 4.2 Proposed DP-ACDKM
- 4.3 Privacy Analysis
- 5 Experiments
- 5.1 Privacy-Utility Trade-Off
- 5.2 Convergence
- 6 Conclusion
- References
- DADR: A Denoising Approach for Dense Retrieval Model Training
- 1 Introduction
- 2 Related Work
- 3 Method
- 3.1 Task Formulation
- 3.2 Denoising Approach Based on Dynamical Weight
- 4 Experiment
- 4.1 Dataset and Metrics
- 4.2 Experiment Settings
- 4.3 Experiment Results
- 5 Conclusions
- References
- Multi-pair Contrastive Learning Based on Same-Timestamp Data Augmentation for Sequential Recommendation
- 1 Introduction
- 2 Related Work
- 2.1 Self-supervised Learning
- 2.2 Sequential Recommendation
- 3 The Proposed Model
- 3.1 Problem Definition
- 3.2 Model Framework
- 3.3 Data Augmentaion
- 3.4 Masking Operation
- 3.5 Embedding Layer
- 3.6 BERT Encoder
- 3.7 Prediction Layer
- 3.8 Multi-pair Contrastive Learning
- 4 Experiments
- 4.1 Experimental Setup
- 4.2 Experimental Results
- 4.3 Hyperparameter Experiments
- 4.4 Ablation Study
- 5 Conclusion
- References
- Enhancing Collaborative Features with Knowledge Graph for Recommendation
- 1 Introduction
- 2 Related Work
- 3 Preliminaries
- 4 Methodology
- 4.1 KG Explore Module
- 4.2 Multi-IMP-GCN
- 4.3 Model Prediction
- 4.4 Model Optimization
- 5 Experiments
- 5.1 Experimental Settings
- 5.2 Performance Comparison
- 5.3 Ablation Studies
- 5.4 Impacts of Multi-IMP-GCN
- 6 Conclusion and Future Work
- References
- PageCNNs: Convolutional Neural Networks for Multi-label Chinese Webpage Classification with Multi-information Fusion
- 1 Introduction
- 2 Multi-label Chinese Webpage Classification Models
- 3 Experimental Results and Discussions
- 3.1 Multi-label Chinese Webpage Dataset
- 3.2 Implementation Details and Evaluation Metrics
- 3.3 Multi-label Chinese Webpage Classification Results
- 4 Conclusion
- References
- MFF-Trans: Multi-level Feature Fusion Transformer for Fine-Grained Visual Classification
- 1 Introduction
- 2 Related Works
- 2.1 CNN-Based FGVC Methods
- 2.2 ViT-Based FGVC Methods
- 3 Proposed Method
- 3.1 Vision Transformer Encoder
- 3.2 Important Token Election Module
- 3.3 Semantic Connection Enhancing Module
- 4 Experiments
- 4.1 DataSets and Implement Details
- 4.2 Comparisons with Advanced Methods
- 4.3 Ablation Studies
- 5 Conclusion
- References
- Summarizing Doctor's Diagnoses and Suggestions from Medical Dialogues
- 1 Introduction
- 2 Related Work
- 3 Model
- 3.1 Pointer Generator Network as Backbone
- 3.2 Input Token Enhancement by Speaker-level Embedding
- 3.3 Input Token Enhancement by Utterance-level Embedding
- 4 Experiment
- 4.1 Dataset
- 4.2 Baseline Models
- 4.3 Settings
- 4.4 Evaluation Metrics
- 4.5 Automatic Evaluation
- 4.6 Doctor Evaluation
- 4.7 Case Study
- 5 Conclusion
- References
- HSA: Hyperbolic Self-attention for Sequential Recommendation
- 1 Introduction
- 2 Preliminaries and Related Work
- 2.1 Empirical Analysis of Datasets
- 2.2 Lorentz Model of Hyperbolic Space
- 2.3 Self-attention Mechanism for Sequential Recommendation
- 3 Proposed Approach
- 3.1 Problem Formulation and Approach Overview
- 3.2 Item Embeddings in Hyperbolic Space
- 3.3 Sequence Learning with Self-attention Mechanism
- 3.4 Prediction Layer
- 3.5 Model Training
- 4 Experiments
- 4.1 Experimental Setup
- 4.2 Experimental Results
- 4.3 Performance as a Plugin on Baselines
- 5 Conclusion
- References
- CFGCon: A Scheme for Accurately Generating Control Flow Graphs of Smart Contracts
- 1 Introduction
- 2 Related Work
- 3 Preliminaries
- 4 Proposed Scheme
- 4.1 Overview of the System Model
- 4.2 Transform Module
- 4.3 Division Module
- 4.4 Connection Module
- 5 Experiment and Performance Evaluation
- 5.1 Dataset
- 5.2 General Test for CFGCon
- 5.3 Performance Comparision with Existing Approaches
- 6 Conclusion
- References
- Hypergraph-Enhanced Self-supervised Heterogeneous Graph Representation Learning
- 1 Introduction
- 2 Related Work
- 2.1 Heterogeneous Graph Embedding
- 2.2 Hypergraph Embedding
- 2.3 Self-supervised Learning on Graphs
- 3 Problem Formulation
- 4 The Design of HHGR
- 4.1 Hypergraph Construction
- 4.2 Hypergraph Encoder
- 4.3 Cross-View Contrast
- 4.4 Semantic Positive Samples
- 5 Experiments
- 5.1 Experimental Settings
- 5.2 Node Classification
- 5.3 Node Clustering
- 5.4 Visualization
- 5.5 Ablation Studies
- 5.6 Analysis of Hyper-parameters
- 6 Conclusion
- References
- LAF: A Local Depth Autoregressive Framework for Cardinality Estimation of Multi-attribute Queries
- 1 Introduction
- 2 Related Work
- 3 Problem Description
- 3.1 Supported Queries
- 3.2 Density Estimation
- 4 Local Deep Autoregressive Framework
- 4.1 Overview
- 4.2 Encoding Tuples and Dividing Subsets
- 4.3 Model Architectures
- 4.4 Model Training with Data and Query
- 4.5 Estimating Cardinality
- 5 Experiment
- 5.1 Experimental Settings
- 5.2 Performance Comparison
- 5.3 Attribute Ordering
- 6 Conclusion
- References
- MGCN-CT: Multi-type Vehicle Fuel Consumption Prediction Based on Module-GCN and Config-Transfer
- 1 Introduction
- 2 Related Work
- 2.1 Fuel Consumption Prediction
- 2.2 Time Series Prediction
- 2.3 Transfer Learning
- 3 Preliminary
- 4 Method
- 4.1 Module Graph with Domain Knowledge
- 4.2 Module Graph Convolution Network
- 4.3 Configuration Transfer Module Based on Classifier
- 4.4 Pre-train and Prediction
- 5 Experiments
- 5.1 Experiments Setup
- 5.2 Fuel Consumption Prediction Performance
- 5.3 Ablation Studies
- 5.4 MGCN Module Performance
- 5.5 CT Module Performance
- 6 Conclusion
- References
- Hardware and Software Co-optimization of Convolutional and Self-attention Combined Model Based on FPGA
- 1 Introduction
- 2 Background
- 2.1 Conv+Transformer
- 2.2 Attention Augmented Convolutional
- 2.3 Hardware Optimization
- 3 Algorithm and Hardware Co-optimization
- 3.1 Quantitative
- 3.2 Overall Architecture
- 3.3 Hardware Layer Design
- 3.4 Softmax Module
- 3.5 LayerNorm Module
- 4 Experiments
- 4.1 Experimental Setup
- 4.2 Channel Ratio Comparison Experiment
- 4.3 Hardware Implementation Results
- 4.4 Multi-platform Comparison
- 5 Conclusion
- References
- FBCA: FPGA-Based Balanced Convolutional Attention Module
- 1 Introduction
- 2 Related Works
- 3 Software Design
- 3.1 Lite Transformer
- 3.2 Quantization
- 4 Hardware Design
- 4.1 The Accelerator Architecture
- 4.2 Matrix Multiplication Module
- 4.3 Convolutional Computing Module
- 4.4 Layer Parallel Optimization
- 5 Experiment
- 5.1 Experimental Setup
- 5.2 Experimental Results
- 5.3 Multi-platform Comparison
- 6 Conclusion
- References
- Multi-level Matching of Natural Language-Based Vehicle Retrieval
- 1 Introduction
- 2 Related Work
- 2.1 Vehicle Retrieval
- 2.2 Vehicle Re-identification
- 2.3 Video Retrieval Based on Natural Language
- 3 Method
- 3.1 Data Enhancement
- 3.2 Representation Learning
- 3.3 Loss Calculation
- 4 Experiment
- 4.1 Settings
- 4.2 Metrics
- 4.3 Comparison Results
- 4.4 Ablation Experiment
- 4.5 Result
- 5 Conclusion
- References
- Improving the Consistency of Semantic Parsing in KBQA Through Knowledge Distillation
- 1 Introduction
- 2 Method
- 2.1 Problem Definition
- 2.2 Teacher Guidance Model
- 2.3 Student Dynamic Distillation Learning Model
- 2.4 Dynamic Weight Assignment Model
- 3 Experiments and Result Analysis
- 3.1 Experimental Settings
- 3.2 Experimental Results
- 4 Related Work
- 5 Conclusion and Future Work
- References
- DYGL: A Unified Benchmark and Library for Dynamic Graph
- 1 Introduction
- 2 The Framework Design
- 2.1 Dataset Structures
- 2.2 Data Normalization Initiative
- 2.3 Comprehensive Models
- 2.4 Fast Initialization and Unified Process
- 3 Experiments with Library
- 3.1 Compared Methods
- 3.2 Datasets
- 3.3 Experimental Setting
- 3.4 Results and Discussion
- 3.5 Results
- 4 Conclusion
- References
- TrieKV: Managing Values After KV Separation to Optimize Scan Performance in LSM-Tree
- 1 Introduction
- 2 Background and Motivation
- 3 TrieKV Design
- 3.1 Architectural Overview
- 3.2 DyTrieIndex
- 3.3 pbCompaction
- 3.4 Main Procedure
- 4 Evaluation
- 4.1 Experimental Setup
- 4.2 Range Query Performance
- 4.3 Point Query Performance
- 4.4 Write Performance
- 4.5 Evaluate TrieKV Under YCSB Default Workloads
- 5 Related Work
- 6 Conclusion
- References
- Bit Splicing Frequent Itemset Mining Algorithm Based on Dynamic Grouping
- 1 Introduction
- 2 Related Work
- 3 Proposed Method
- 3.1 Preprocessing
- 3.2 Dynamic Grouping
- 3.3 Build MPL
- 3.4 Mining Based on MPL
- 4 Pruning Strategy
- 4.1 Consecutive Supersets Pruning
- 4.2 Grouping Pruning
- 5 Experiment
- 5.1 Test Environment and Dataset
- 5.2 Time Evaluation
- 5.3 Memory Evaluation
- 5.4 Result Analysis
- 6 Conclusion
- References
- Entity Resolution Based on Pre-trained Language Models with Two Attentions
- 1 Introduction
- 2 Related Work
- 3 Pre-trained Model with Two Attentions
- 3.1 Problem Definition
- 3.2 Framework of IGaBERT Model
- 3.3 Interactive Attention
- 3.4 Comparison
- 3.5 Global Attention
- 3.6 Gate and Binary Classifier
- 4 Experimental Results
- 4.1 Parameters and Settings
- 4.2 Data Augmentation
- 4.3 Datasets
- 4.4 Comparison with Existing Models
- 4.5 Ablation Study
- 4.6 Case Study
- 5 Conclusion
- References
- A High-Performance Hybrid Index Framework Supporting Inserts for Static Learned Indexes
- 1 Introduction
- 2 Motivation
- 2.1 Dynamic PGM-Index
- 2.2 The Problem
- 2.3 The Reason
- 3 Hybrid Index Framework
- 3.1 Overview
- 3.2 The Dynamic Layer for Inserts
- 3.3 The Static Layer for Lookups
- 3.4 Self-tuning Algorithm
- 4 Performance Evaluation
- 4.1 Experimental Settings
- 4.2 Performance Evaluation Under Read-Write Workloads
- 4.3 Lookup Performance Under Read-Write Workloads
- 4.4 Insert Performance Under Read-Write Workloads
- 4.5 Learning Overhead of Hybrid Index Framework
- 5 Related Work
- 6 Conclusion
- References
- A Study on Historical Behaviour Enabled Insider Threat Prediction
- 1 Introduction
- 2 Related Works
- 3 Methodology
- 3.1 Problem Setting
- 3.2 Framework
- 3.3 ML and DL Models
- 4 Experiments
- 4.1 Dataset and Data Processing
- 4.2 ML and DL Performance Comparison
- 4.3 The Impact of Historical Window Length
- 5 Conclusion
- References
- PV-PATE: An Improved PATE for Deep Learning with Differential Privacy in Trusted Industrial Data Matrix
- 1 Introduction
- 2 Related Work
- 2.1 DP Mechanisms for DL
- 2.2 DP-SGD vs PATE
- 3 Background and Notation
- 3.1 Differential Privacy
- 3.2 Rényi Differential Privacy
- 3.3 PATE and PV-PATE
- 4 Personalized Voting for PATE in TDM
- 4.1 Teacher Credibility
- 4.2 Adaptive Voting Mechanism
- 4.3 Model Sharing Mechanism
- 4.4 Privacy Analysis
- 5 Experiments
- 5.1 Datasets and Models
- 5.2 Aggregation Mechanism with Teacher Credibility
- 5.3 Student Training with Adaptive Voting Mechanism
- 5.4 Student Training with Model Sharing Mechanism
- 6 Conclusion
- References
- LayerBF: A Space Allocation Policy for Bloom Filter in LSM-Tree
- 1 Introduction
- 2 Background and Motivation
- 2.1 LSM-Tree
- 2.2 Bloom Filter
- 2.3 Motivation
- 3 Design
- 3.1 Main Idea
- 3.2 Minimize Average False Positive Rate
- 3.3 LayerBF Allocation Strategy
- 4 Evalution
- 4.1 Experimental Setup
- 4.2 Average False Positive Rate
- 4.3 Overall Performance Evaluation
- 4.4 YCSB Benchmarks
- 5 Related Work
- 6 Conclusion
- References
- HTStore: A High-Performance Mixed Index Based Key-Value Store for Update-Intensive Workloads
- 1 Introduction
- 2 Background and Motivation
- 2.1 Log-Structured Merge-Tree (LSM-Tree)
- 2.2 Non-volatile Memory (NVM)
- 2.3 Motivation
- 3 HTStore Design
- 3.1 Mixed Index Design
- 3.2 Eliminating Redundant Data
- 3.3 Accelerating Point Query
- 4 Performance Evaluation
- 5 Related Work
- 6 Conclusions and Future Work
- References
- Author Index
System requirements
File format: PDF
Copy protection: Watermark-DRM (Digital Rights Management)
System requirements:
- Computer (Windows; MacOS X; Linux): Use the free software Adobe Reader, Adobe Digital Editions, or any other PDF viewer of your choice (see eBook Help).
- Tablet/Smartphone (Android; iOS): Install the free app Adobe Digital Editions or another reading app for eBooks, e.g., PocketBook (see eBook Help).
- E-reader: Bookeen, Kobo, Pocketbook, Sony, Tolino and many more (only limited: Kindle).
The file format PDF always displays a book page identically on any hardware. This makes PDF suitable for complex layouts such as those used in textbooks and reference books (images, tables, columns, footnotes). Unfortunately, on the small screens of e-readers or smartphones, PDFs are rather annoying, requiring too much scrolling.
This eBook uses Watermark-DRM, a „soft” copy protection. This means that there are no technical restrictions to prevent illegal distribution. However, there is a personalised watermark embedded in the eBook that can be used to identify the purchaser of the eBook in the event of misuse and to provide evidence for legal purposes.
For more information, see our eBook Help page.