Graph Analytics for Big Data

An introduction to the course where you meet the instructor, Amarnath Gupta, and learn about the course objectives. This module provides a brief welcome and overview of what to expect in the Graph Analytics for Big Data course.

This module provides a first exposure to graphs and their use in everyday life. It covers the definition of graphs, why graphs are useful for modeling various systems, and explores examples including social networking, biological networks, human information network analytics, and smart cities. The module also discusses the purpose of analytics and the impact of Big Data's V's on graphs.

This module focuses on graph analytics techniques including path analytics, connectivity analytics, community analytics, and centrality analytics. It covers Dijkstra's algorithm, inclusion and exclusion constraints, graph disconnection, indegree and outdegree, local and global properties, modularity, and various centrality measures. Optional lectures include bi-directional Dijkstra algorithm, goal-directed Dijkstra algorithm, power law graphs, graph evolution measurement, eigenvector centrality, and key player problems.

This module provides hands-on experience with Neo4j, a free and powerful graph analytics tool. It covers how to use Cypher, the query language of Neo4j, to perform a wide range of analyses on graph networks. Topics include getting started with Neo4j, modifying graphs, importing data, basic queries, path analytics, and connectivity analytics using Cypher. The module includes practical demonstrations and exercises to apply graph analytics techniques.

This final module examines how graph analytics and graph data management come together through specialized programming models and software frameworks. It introduces parallel programming models for graphs, the Pregel system, and frameworks like Giraph and GraphX. The module includes hands-on demonstrations with GraphX, covering building a graph, creating degree histograms, plotting histograms, analyzing network connectedness, clustering components, and joining graph datasets.