CS 6501: Empirical Software Engineering , Spring 2017

Lecture Details

Instructor: Baishakhi Ray (Rice 340)
Meetings : Tuesday and Thursday 3:30 PM - 4:45 PM at Rice Hall 340
Office Hours: Tuesday 5 pm - 6 pm or by appointment
Teaching Assistant: Chong Tang (ctang at virginia dot edu)

Description

This course will teach how to write better software using diverse data analysis techniques borrowed from Machine Learning, Natural Language Processing, Network Analysis, and Statistical Modeling. We will analyze large-scale open-source software data from GitHub and StackOverflow, and build different tools such as code recommendation systems, automatic bug finders, etc. In addition, we will read research papers to learn the current start-of-the-art.

Prerequisite

The class will primarily focus on Software Engineering Research. You should have a good knowledge about program development and handling moderate to large scale code repository. It will require at least 8-10 hours + class time per week for this class.

Tentative Topics

• Program Variants
• Software Enginnering + Natural Language Processing (NLP)
• Software Enginnering + Information Retrieval (IR)
• Mining Software Repositories (MSR)

Schedule

Reading & Presentation	Lectures	Assignments	Presenter
Introduction	Week 1 (1/19/2017)	Overview	Baishakhi Ray
Program Variants	Week 2 (1/24/2017)	Program Tuning	Chong Tang
	Week 2 (1/26/2017)	Cost of maintaining program variant A Case Study of Cross-System Porting in Forked Software Projects Detecting and Characterizing Semantic Inconsistencies in Ported Code	Baishakhi Ray

Grading

• Group Project (2-3 students) – 60%
• Project Proposal - 5%
• Mid-term report/presentation - 15%
• End of semester presentations/demo – 20%
• End of semester project report - 20%
• Quiz – 25%
• Class Participation – 15%
  • Paper presentation - 8%
  • Reviews/Questions - 7%