Overview

This course introduces the basics of parallel and distributed computing with an experimental focus on online social network analysis. The analysis of online social networks has been chosen as the target class of applications for applying the knowledge introduced in lectures because of several factors: it is an active research topic in computer science; the scale challenges of very large social networks require parallel computing approaches; and parallelizing graph algorithms is not trivial.

The course is project-oriented. Class time will be spent on a combination of lectures on topics in parallel programming and models, in-class discussions of research papers, mini-lectures on Social Network Analysis (SNA), brainstorming sessions on parallelizing some SNA-relevant algorithms, and interdisciplinary team project meetings.  There are no tests. Final evaluation will be based on class participation (20%), programming projects (40%), and final project and presentation (40%). Lectures, readings and class discussions will introduce the design of parallel programs, performance metrics, distributed systems architectures, time and global state, coordination and agreement, distributed transactions and web services. Programming projects will introduce the practice of parallel and distributed programming with threads, MPI, Google’s MapReduce framework, and Amazon Web Services. The final project will be an interdisciplinary team project that involves graduate students enrolled in the Social Network Analysis course (SYA 6933.904) taught by Professor Skvoretz from the Department of Sociology. 

The course is intended for advanced undergraduate students and graduate students in CSE.

Schedule

Final Projects