Network Analysis
Why?
Social Network Analysis is essential for understanding how relationships and connections shape our digital world. From analyzing social media networks to understanding knowledge graphs, this course gives you the mathematical foundation and practical tools to analyze complex networks. You'll learn the algorithms behind Google's PageRank, Facebook's friend recommendations, and how companies like LinkedIn analyze professional networks. It's where graph theory meets real-world applications in data science.
What?
A course that bridges graph theory mathematics with real-world knowledge graph applications. You'll start with fundamental graph concepts (vertices, edges, adjacency matrices) and graph theory, then move into network analysis using centrality measures to identify important nodes, then take an introduction to machine learning on graphs, while you take a glimpse into the world of knowledge graphs.
Curriculum:
Introduction to Knowledge Graphs
Understanding what knowledge graphs are and why they matter. Covers semantic descriptions of entities and relationships, the rise of knowledge graphs in industry (Google, Facebook, LinkedIn Economic Graph), and how they enable intelligent applications like semantic search and question answering.
Graph Theory Fundamentals
Core concepts: vertices (nodes/points) and edges (links/arcs), directed vs undirected graphs, graph size notation (n=|V|, m=|E|), neighborhoods, graph classes including cycle graphs, complete graphs, and bipartite graphs. Graph traversal concepts: walk, path, circuit, and cycle definitions with examples.
Graph Representation Methods
Three main ways to represent graphs: Adjacency Matrix A(G), Incidence Matrix, and Adjacency List. Understanding when to use each method, their space complexity, and how to work with weighted graphs. Includes practical examples and matrix calculations.
Spanning Trees & Network Algorithms
Spanning tree definitions and properties, minimum spanning tree algorithms. Connected vs disconnected graphs, graph connectivity analysis, and basic network algorithms for finding shortest paths and analyzing graph structure.
Network Centrality Analysis
Key centrality measures: Degree Centrality, Closeness Centrality, Betweenness Centrality, and Eigenvector Centrality. Learning to calculate each measure, understand normalization formulas, and interpret results to identify important nodes in networks.
Real-World Knowledge Graph Applications
Case studies: Google Knowledge Graph for search enhancement, Wikidata for open knowledge, LinkedIn Economic Graph, ResearchSpace for cultural heritage, and life sciences applications. Understanding how knowledge graphs solve real business problems.
Machine Learning for Graph Predictions
Three types of graph predictions: Node-level (classification), Link-level (predicting new connections), and Graph-level (properties of entire graphs). Feature engineering for graphs, kernel methods, graphlets, and introduction to the ML pipeline for graphs.
