CS846: Topics in Empirical Software Evolution

• The beta-draft of website for W23 CS846 is now up and running; please let me know about any mistakes, inconsistencies you find or any questions you have.
• Classes will be held in person in DC2585 from 2:30-5:30 pm

• On a typical week, there will be two or three paper presentations and discussions led by students; each student will likely end up doing one or two presentations over the term
• All students are expected to come to class having read all of the chapters/papers for the week and be prepared to engage in a discussion about them
• Each week, a short written review of the research papers (but not the book chapters) must be submitted by each student
• There will be five parts to your grade:
  1. in-class discussion participation (15%)
  2. weekly paper reviews (15%)
  3. one or two paper presentations (10%)
  4. a project done in groups of up to 3 students (50%)
  • 5% for the project proposal and its presentation
  • 45% for the final project report and its presentation
  5. participation in reviewing and discussing two other class projects (10%)
• The course project could be a topic of your own choosing, related to the course themes, or you could decide to do the 2023 Mining Software Repositories Challenge; you can pick which of the questions most interest you, or devise your own using the provided dataset. Note that you are not required to submit your project to the conference itself (the deadline is in early February).
• A little after the mid-way point in the term, each project group will give a formal 5–10 minute presentation of their project proposal to the class, who will in turn provide critical feedback
• After the final projects have been submitted, each student be expected to review two other projects and engage in an online discussion with other anonymous class mates, using a conference submission management system such as EasyChair (we will set this up for you).
  • Authors of the project being reviewed will (eventually) receive a unified meta-review, but will not be able to see the initial reviews or discussions.
  • The review process will be double-blind: The project authors will not know the identities of those taking part in reviewing their project; also, reviewers will not know the identities of their fellow reviewers. I will monitor the reviews (and I will be the only person who knows everyone's identify. The purpose of this is to give you all some experience in reviewing research.
  • The reviews are meant to be helpful feedback to the project authors; they will not serve as input into the assigning of the grades.