COMP 6611B: Topics on Cloud Computing and Data Analytics Systems [Fall 2015]

Course Format

This is a seminar style course where each week we will cover a group of papers focusing on a specific topic of cloud and data analytics systems. Students are expected to read all the papers that will be presented in class and submit a review for one of them before attending the class. Students will be given papers to present in class and will work on a term-long, open-ended course project.

Paper Review

A paper review should strike the following points.

  1. Paper summary: Briefly summarize the major issues addressed in the paper. Do you consider them important? You should also comment on the novelty, creativity, contribution, impact, and technical depth in the paper.

  2. Strengths: Briefly list some bulleted points of the strengths of the paper. What are the major reasons that make you like the paper?

  3. Weaknesses: Briefly list some bulleted points of the weaknesses of the paper. What are the major issues that you believe have not been well addressed in the paper? E.g., model assumptions, solution practicality, implementation complexity, evaluation completeness, etc.

  4. Detailed comments: You should provide detailed comments, explanations, and justifications on the strengths and weaknesses summarized above. For the weaknesses, you are highly encouraged to suggest promising alternatives that may address the problems.

Class Presentation

Each student will present at least one paper in the entire course. (The number of presentations depends on the number of students enrolled in the course.) At the beginning of each week, we will determine the presenters and the papers that will be presented in the Friday lecture and the Monday lecture in the following week. Each presentation is limited to 25 minutes. After the presentation, we will randomly choose students to ask/answer questions and discuss the important issues addressed in the paper. We will also comment on the presentation and paper reviews.

Course Project

In addition to paper reviews and class presentations, the course consists of a term-long, open-ended project. The topic of the project is mainly determined by you, but must be approved by the instructor. Example topics include but not limited to

  • Measuring, analyzing, and characterizing the cluster workload traces published by Google;

  • Implementing, measuring, and comparing resource management policies proposed in the existing literature;

  • Reproducing the simulation/experimental results reported in the papers that have been covered in the classes;

  • An in-depth literature survey focusing on a specific topic covered in the course (the surveyed papers are not limited to the reading list);

  • Any new research ideas and preliminary results that are closely related to the course topics.

  • ...

The delivery of the project is a 6-page course thesis (not including references) along with a 10-minute presentation. The presentation will be marked by the instructor and the audiences. You could either work on the project by you own, or collaborate with another student. If you choose to collaborate, the final presentation will be up to 15 minutes. You are required to submit a one-page project proposal at the beginning of the term, a 3-page mid-term report, and the final thesis at the end of the term. All the documents must be PDF files produced by LaTeX, using the ACM conference template.