CS 521/621: Advanced Software Engineering: Analysis and Evaluation

Fall 2013

News | Description | Logistics | Grading | Schedule | Academic integrity | Reading | Prerequisites | Acknowledgements



Software engineering goes well beyond software development. It involves understanding customer needs, planning the development process, teamwork, maintenance, and analysis and evaluation tasks, such as testing, verification, and validation. In CS 521/621, we will explore the state-of-the-art research in techniques that make analysis and evaluation of software easier. Concurrently, student groups will engage in research to push forward that state of the art.

This term will cover the following topics: static, dynamic, and speculative analyses; model inference, model checking, and formal verification; bug localization; mutation and regression testing, and symbolic execution.

Becoming familiar, and a part of cutting-edge research will constitute a major part of this class. There will be three research aspects of the class:

  1. The lectures and assignments will draw on the latest research in software engineering
  2. The students will, either individually or in pairs, engage in reading and presenting to the class existing research from recent premier conferences and journals. The students will summarize the work in writing and in a class presentation, and will lead a class discussion.
  3. The students, either individually or in groups of up to four, will engage in a term-long project to advance the state of the art of one of the above topics. The students will identify a shortcoming of today's techniques, review the relevant literature, develop a novel technique that addresses the shortcoming, and evaluate the technique against the state of the art. Finally, the goal is for students to write a research paper describing their work worthy of submission to a research conference.

The 621 students will be required to present an existing research paper once during the semester, and will complete the research project. Meanwhile, the 521 students will present an existing research paper twice during the semester, or may opt to complete a research project for extra credit, and in lieu of one of the presentations.


Room:142 computer science building
Lecture:Monday and Wednesday 10:35AM–11:50AM

Yuriy Brun
office: 346 computer science building
office hours: Wednesday 12:00PM–1:00PM

Andrew Kae
office: CS 256 (computer science building)
office hours: Tuesday 3:00PM–4:00PM

All assignment submissions are through Moodle.

Late policy: Assignment due dates and times are listed on the schedule. All deadlines are sharp and the submission site will be closed at the specified time. No extensions will be granted after the assignment is due. Early requests for extensions will be considered only in extenuating circumstances. No more than one extension will be granted per student.


Students are responsible for submitting all homework and project assignments. A student who fails to submit at least one of the homework or project assignments, or does not participate in the midterm, will receive the grade F for the entire class.

 Assignment Grade
Midterm 30%
Homework 35%
Paper summary and presentations 30%
Participation 5%

 Assignment Grade
Midterm 20%
Homework 35%
Paper summary and presentation 15%
Participation 5%
Project 25%

The project's 25% are further broken down:

literature review 7%
project plan and presentation 7%
final report and presentation 11%
total: 25%


(subject to change; check regularly)

Due: M Nov 25, 2013, 9:00AM EST
week date day topic reading homework
Week 1
Sep 4 W Course Introduction
Week 2
Sep 9 M Static and Dynamic Analyses Overview
Sep 11 W Models, Tests, Bugs, and Symbols Overview
Homework 1
Due: M Sept 23, 2013, 9:00AM EDT
Week 3
Sep 16 M Dynamic Analysis Dynamically discovering likely program invariants to support program evolution & Purify
Paper Selection and Idea Proposal
Due: M Sept 30, 2013, 9:00AM EDT
Sep 18 W Automated Test Generation, and
"On the Naturalness of Software" by Prem Devanbu (watch video)
Korat: Automated testing based on Java predicates &
On the Naturalness of Software
Week 4
Sep 23 M Reproducing Field Failures Chronicler: Lightweight Recording to Reproduce Field Failures
Sep 25 W Pair Programming
Homework 2
Due: M Oct 7, 2013, 9:00AM EDT
Week 5
Sep 30 M Research idea presentations
Oct 2 W Speculative Analysis Early Detection of Collaboration Conflicts and Risks
Literature review
Due: Tu Oct 21, 2013, 9:00AM EDT
Week 6
Oct 7 M Performance debugging
Oct 9 W no class. Instead, attend talk on Oct 10.
Homework 3
Due: W Oct 23, 2013, 9:00AM EDT
Oct 10 R "Data Provenance and Query Provisioning" by Val Tannen (watch video)
Week 7
Oct 14 M Bartolomé day: no class
Oct 15 T Monday schedule: Design exercise
Oct 16 W Design exercise
Week 8
Oct 21 M Paper presentations on automated bug repair Automatic Patch Generation Learned from Human-Written Patches, Automatic Recovery from Runtime Failures, and Modular and Verified Automatic Program Repair.
Oct 23 W Paper Presentations on testing Using likely invariants for automated software fault localization, Finding your way in the testing jungle: A learning approach to web security testing, and Automated testing with targeted event sequence generation.
Project plan
Due: M Nov 4, 2013, 9:00AM EST
Week 9
Oct 28 M Privacy and Reliability Entrusting Private Computation and Data to Untrusted Networks, and
Smart redundancy for distributed computation
Homework 4
Due: W Nov 13, 2013, 9:00AM EST
Oct 30 W Paper presentations on defect prediction and fixing SemFix: Program Repair via Semantic Analysis, Data Clone Detection and Visualization in Spreadsheets, and Sample Size vs. Bias in Defect Prediction.
Week 10
Nov 4 M "Architectural Decay in Software Systems: Symptoms, Causes, and Remedies" by Nenad Medvidovic (watch video)
Nov 6 W Paper presentations on refactoring RefaFlex: Safer Refactorings for Reflective Java Programs, Refactoring with Synthesis, and Drag-and-Drop Refactoring: Intuitive and Efficient Program Transformation.
Final project report
Due: Fr Dec 6, 2013, 11:59PM EST
Week 11
Nov 11 M Veterans day: no class
Nov 13 W Midterm review and Software Architecture
Week 12
Nov 18 M Midterm
Homework 5
Due: W Nov 27, 2013, 9:00AM EST
Nov 20 W Project work time: no class
Week 13
Nov 25 M Paper presentations on regression logs and tests
Be Conservative: Enhancing Failure Diagnosis with Proactive Logging, Will You Still Compile Me Tomorrow? Static Cross-Version Compiler Validation, and Guided Test Generation for Web Applications.
Nov 27 W Paper presentations on verification Patterns in Property Specifications for Finite State Verification and Injecting Mechanical Faults to Localize Developer Faults for Evolving Software.
Week 14
Dec 2 M Project presentations
Dec 4 W Zero Knowledge Proofs and more

Academic integrity:

Students are allowed to work together on all aspects of this class except the midterm. However, for the homework assignments, each student must submit his or her own write up, clearly stating the collaborators. Your submission must be your own. When in doubt, contact the instructors about whether a potential action would be considered plagiarism. If you discuss material with anyone besides the class staff, acknowledge your collaborators in your write-up. If you obtain a key insight with help (e.g., through library work or a friend), acknowledge your source and write up the summary on your own. It is the student's responsibility to remove any possibility of someone else's work from being misconstrued as the student's. Never misrepresent someone else's work as your own. It must be absolutely clear what material is your original work. Plagiarism and other anti-intellectual behavior will be dealt with severely. Note that facilitation of plagiarism (giving your work to someone else) is also considered to be plagiarism, and will carry the same repercussions.

Students are encouraged to use the Internet, literature, and other publicly-available resources, except the homework solutions and test (including quizzes, midterms, finals, and other exams) solutions, from past terms' versions of this course and other academic courses, whether at UMass and at other institutions. To reiterate, the students are not allowed to view and use past homework and test solutions, unless explicitly distributed by the CMPCSI 521/621 staff as study material.

Whenever students use Internet, literature, and other publicly-available resources, they must clearly reference the materials in their write ups, attributing proper credit. This cannot be emphasized enough: attribute proper credit to your sources. Failure to do so will result in a zero grade for the assignment and possibly a failing grade for the class, at the instructor's discretion. Copying directly from resources is not permitted, unless the copying is clearly identified as a quote from a source. Most use of references should be written in the words of the student, placing the related work in proper context and describing the relevant comparison.

The students should familiarize themselves with the UMass Academic Honesty Policy and Guidelines for Classroom Civility and Respect. These policies and guidelines apply to this class.

Students who violate University standards of academic integrity are subject to disciplinary sanctions, including failure in the course and suspension from the university. Since dishonesty in any form harms the individual, other students, and the university, policies on academic integrity have been and will be strictly enforced.


There is no required textbook for the course. Reading assignments will come from publicly available research papers. Students who wish to read established textbooks beyond the assigned reading should consider:


Students should have taken an introductory course in software engineering or have the equivalent background. Students are expected to be familiar with an object oriented programming language, such as Java or C++. Some programming and the ability to download and use off-the-shelf tools are expected.


Various materials used in this course have greatly benefited from materials developed by Alex Aiken, Lori Clarke, Carlo Curino, Sebastian Elbaum, Michael Ernst, David Notkin, Nenad Medvidovic, Alex Orso, Lee Osterweil, Willem Visser. Thank you.