Applied Graph Data Science Graph Algorithms and Platforms, Knowledge Graphs, Neural Networks, and Applied Use Cases

• Main Authors: Raj, Pethuru, Dutta, Pushan Kumar, Chong, Peter Han Joo, Song, Houbing Herbert, Zaitsev, Dmitry A
• Format: eBook
• Language: English
• Published: Chantilly Elsevier Science & Technology 2025
• Edition: 1
• ISBN: 9780443296543; 0443296545
• DOI: 10.1016/C2023-0-50454-9

Applied Graph Data Science: Graph Algorithms and Platforms, Knowledge Graphs, Neural Networks, and Applied Use Cases delineates how graph data science significantly empowers the application of data science.The book discusses the emerging paradigm of graph data science in detail along with its practi...

Table of Contents:

• 4.1 Graph neural networks for temporal knowledge representation -- 4.2 Experimental results and comparison with baseline methods -- 5. ML-based chronological event detection -- 5.1 Feature engineering for event recognition -- 5.2 Machine learning models for event detection -- 5.3 Performance evaluation and case studies -- 6. Temporal link prediction using ML -- 6.1 Modeling temporal relationships for link prediction -- 6.2 Quantitative analysis and evaluation metrics -- 7. Applications and case studies -- 7.1 Predictive analysis of historical trends -- 7.2 Real-world applications of AI-driven chronological reasoning -- 7.2.1 Environmental monitoring -- 7.2.2 Healthcare and disease outbreak prediction -- 7.2.3 Manufacturing and quality control -- 7.2.4 Supply chain optimization -- 7.2.5 Customer behavior analysis -- 7.2.6 Personalized recommendations in E-commerce -- 7.2.7 Social network analysis -- 7.2.8 Urban planning and traffic management -- 8. Advantages and limitations of the proposed framework -- 8.1 Advantages -- 8.2 Limitations -- 9. Future directions -- 10. Conclusion -- Further reading -- 3 - Graph based approach on financial fraudulent detection and prediction -- 1. Introduction -- 2. Classification of frauds -- 3. Graph network and research direction -- 3.1 Simple versus bipartite -- 3.2 Homogeneous versus heterogeneous -- 3.3 Directed versus undirected -- 3.4 Static versus dynamic -- 3.5 Attributed versus unattributed -- 4. Different time-domain keeping track of network user activities over different timestamps -- 5. Multiplex behavior of users and network support -- 6. Eliminating human intervention in structural information extraction -- 7. Financial fraud and graphical neural network -- 7.1 Approaches of graphical neural network -- 7.2 Different types of graphs -- 8. Graph attention network -- 8.1 Attention score
• 1.1 Limitations of traditional keyword-based approaches -- 1.2 Rationale for integration of graphs, language models, and NLP -- 1.2.1 Growing complexity of user queries -- 1.2.1.1 Evolving internet landscape -- 2. Graphs in search engines -- 2.1 Introduction to graph-based approaches -- 2.1.1 Definition of graphs in the context of search -- 2.1.2 Historical perspective on graphs in information retrieval -- 2.2 Advantages of graph-based representations -- 2.2.1 Enhanced contextual understanding -- 2.2.2 Improved relationship modeling -- 2.3 Challenges and considerations -- 2.3.1 Scalability issues -- 2.3.2 Ethical and privacy concerns -- 3. Language models and NLP in search -- 3.1 Language models in information retrieval -- 3.1.1 Introduction to language models -- 3.1.2 Applications in search engines -- 3.2 NLP techniques for enhanced search -- 3.2.1 Semantic search -- 3.2.2 Entity recognition and contextual understanding -- 3.3 Personalized search with language models -- 3.3.1 Tailoring results to user preferences -- 3.3.2 Dynamic language generation -- 4. Intersection of graphs, language models, and NLP -- 4.1 Integrating language models into graph structures -- 4.1.1 Creating interconnected nodes -- 4.1.2 Case studies demonstrating effectiveness -- 4.2 NLP techniques for graph-based search -- 4.2.1 Multilingual search capabilities -- 4.2.2 Addressing complex user queries -- 5. Case studies and applications -- 5.1 Google's knowledge graph -- 5.1.1 Overview and impact on search results -- 5.1.2 Integration with language models -- 5.1.3 Impact on user experience -- 5.1.4 Challenges and considerations -- 5.2 Open AI's GPT-4 in search engines -- 5.2.1 Utilizing advanced language models -- 5.2.2 Improving search relevance -- 6. Challenges and future directions -- 6.1 Ethical considerations in graph-based search -- 6.1.1 Responsible use of user data
• 3.2 Interpretability of GNNs -- 4. Applications of graph neural networks -- 5. Open problems and future research directions -- 6. Conclusion -- References -- 5 - Delineating graph neural networks (GNNs) and the real-world applications -- 1. Introduction -- 2. Briefing the distinctions of graphs -- 3. The emergence of graph data science -- 4. The emergence of graph neural networks (GNNs) -- 5. Graph analysis techniques -- 6. Demystifying deep neural networks on graph data -- 7. Graph neural networks (GNNs): The applications -- 8. A case study: Graph analytics for e-commerce -- 9. Types of graph neural networks (GNNs) -- 10. Graph convolutional network (GCN) -- 11. The challenges for graph neural networks (GNNs) -- 12. Conclusion -- 6 - Graph techniques for enhancing knowledge graph integration: A comprehensive study and applications -- 1. Introduction -- 1.1 Motivational example -- 2. Background -- 2.1 Semantic technologies, ontologies, knowledge graph -- 2.1.1 Ontologies -- 2.1.2 Knowledge Graph -- 3. Knowledge graph integration -- 3.1 Matching problem -- 3.2 Terminology used in knowledge graph matching -- 3.3 General process of knowledge graph matching -- 3.4 Heterogeneity and similarity measures -- 4. Graph matching techniques -- 5. Knowledge graph matching methods -- 5.1 Graph based techniques for ontology and knowledge graph matching -- 5.2 Taxonomy-based techniques -- 5.3 Graph neural networks base knowledge graph matching -- 5.4 Intersection of Knowledge Graph Integration, Knowledge Graph Matching, and Graph Matching -- 6. Case study: Enhancing semantic segmentation for remote sensing through knowledge graph matching -- 7. Application perspective of knowledge base integration -- 8. Conclusion -- References -- 7 - Graphs, language models, and NLP: The future of search engines -- 1. Introduction
• 9. Recurrent network -- 10. Conclusion and future direction -- References -- 4 - The power of graph neural networks: From theory to application -- 1. Introduction -- 1.1 Importance of graphs in various domains -- 1.2 Challenges in processing graph-structured data -- 1.3 Motivation for using graph neural networks (GNNs) -- 2. Fundamentals of graph neural networks -- 2.1 Introduction to graph theory concepts -- 2.2 Overview of the design pipeline of GNNs -- 2.2.1 Data representation in graphs -- 2.2.2 Feature engineering for graph data -- 2.2.3 GNNs' pipeline design in a nutshell -- 2.3 Variants of graph neural networks -- 2.3.1 Graph convolutional networks -- 2.3.1.1 Architecture and working principle -- 2.3.1.2 Challenges and possible alternatives -- 2.3.2 Graph recurrent neural networks -- 2.3.2.1 Architecture and working principle of GRNNs -- 2.3.2.2 Challenges and possible alternatives -- 2.3.3 Graph attention networks -- 2.3.3.1 Architecture and working principle of GANs -- 2.3.3.2 Challenges and possible alternatives -- 2.3.4 Graph generative networks -- 2.3.4.1 Architecture and working principle -- 2.3.4.2 The challenges and possible alternatives -- 2.3.4.2.1 Challenges -- 2.3.4.2.2 Possible alternatives -- 2.3.4.3 Additional considerations -- 2.3.5 Spatial-temporal graph neural networks -- 2.3.5.1 Architecture and working principle -- 2.3.5.2 Challenges and possible alternatives -- 2.3.5.2.1 Challenges -- 2.3.5.2.2 Possible alternatives -- 2.3.5.3 Additional consideration -- 2.3.6 Hybrid forms -- 2.3.6.1 Architecture and working principle -- 2.3.6.2 Working principle -- 2.3.6.3 Challenges and possible alternatives -- 2.3.6.3.1 Challenges -- 2.3.6.3.2 Possible alternatives -- 2.3.6.4 Additional consideration -- 2.4 GNNs taxonomy -- 3. Scalability and interpretability of graph neural networks -- 3.1 Scalability of GNNs
• 6.1.2 Ensuring fair and unbiased results
• Front Cover -- Applied Graph Data Science -- Applied Graph Data Science -- Copyright -- Contents -- Contributors -- 1 - Introduction to graph neural network: A systematic review of trends, methods, and applications -- 1. Introduction -- 2. Historical background of GNN -- 3. Architecture of GNN -- 4. Relationship between traditional graph embedding &amp -- GNN -- 4.1 Link prediction and graph categorization -- 4.2 Matching graphs and learning how to structure graphs -- 4.3 Automated machine learning and self-supervised learning -- 5. How GNN is better than CNN -- 6. Difficulties in graph learning process -- 7. Current trends in GNN -- 7.1 Graph Attention Networks -- 7.2 Graph representation learning -- 7.3 Transformer based GNN -- 7.4 Hierarchical GNNs -- 7.5 Classical Graph Generative Models -- 7.5.1 Erdos-Renyi model -- 7.5.2 Barabási Albert model -- 7.5.3 Watts Strogatz model -- 7.5.4 Forest Fire model -- 7.5.5 Stochastic Block Model -- 7.5.6 Exponential Random Graph Models -- 7.6 Graph pooling -- 8. Real world applications of GNN -- 8.1 Recommender systems -- 8.2 Computer vision -- 8.3 Natural Language Processing -- 8.4 Drug development -- 8.5 Social network analysis -- 8.6 Robotics industry -- 8.7 Fraud detection -- 9. Future directions in GNN -- 10. Conclusion -- References -- 2 - Chronological reasoning in knowledge graphs using AI and ML: A novel framework -- 1. Introduction -- 1.1 Background and motivation -- 1.2 Significance of chronological reasoning in knowledge graphs -- 2. Related work -- 2.1 Overview of knowledge graphs and temporal reasoning -- 2.2 Previous approaches to AI and ML in knowledge graph analysis -- 3. Methodology -- 3.1 Data preprocessing and temporal annotation -- 3.2 Temporal embeddings using AI-enhanced ML techniques -- 3.3 ML-based chronological event detection -- 4. AI-enhanced temporal embeddings