Written Analyses are due 5 hours BEFORE class starts

If you choose to do more than the minimum required, the additional analyses will be graded and the lowest homework grade of this type will be dropped.

Details on the written analyses and the presentation requirements are given below.

There is also a new level of flexibility for these assignments. For any given lecture we will have between 2 and 5 papers that will be covered. There may also be additional papers listed as "optional" which are provided only as additional information for the highly motivated.

You will have the flexibility to analyze and write about all, some, or only one of the papers listed for a particular technical topic. The expectation is that you will spend the same amount of time on a given homework assignment regardless of how many of the papers you choose to analyze. If you analyze more papers, the analysis of each will necessarily be less. If you analyze fewer papers, the analysis will necessarily be much deeper. (NOTE: time spent reading and understanding the paper doesn't count since you have to read and understand all of the required papers anyway.)

I've not tried this last twist before, so I'm not sure how exactly it will work, but it will be an interesting experiment. In thinking about this for myself, I'm not sure if I were in your place which I would make. My sense is that if I had a lot to say about a particular paper--that it generated some strong reaction--at least for that assignment I would address fewer papers (or only the one). If I didn't have strong reactions about any of the papers, I would write about each. My hope is that I'm giving you the flexibility to write about the papers for which you have the strongest reaction.

As for guidelines for what you should write as an analysis for a given paper, there are not specific guidelines. As mentioned above, we've talked quite a bit in class about organization. HW02 was a first attempt to look at a paper more deeply (see the list of HW02 submissions and their associated grades for some good examples). And finally, look at the analysis I wrote for the Infocom paper (which I realize I didn't cover much in class, but it really wasn't offered as a technical paper).

Since you will have multiple opportunities to write analyses, pay close attention to the grade and feedback I'll provide. Hopefully over the course of the quarter, everyone will improve.

Similar to the earliest homework assignments, email a copy of
what you write to "almeroth@cs.ucsb.edu" with the Subject line:
"CS290F HW Turnin". *Your name should not appear anywhere on
a printed version of the assignment.* (I will figure out who you
are based on your email address.) Assignments are due 5 hours
before class starts on the day that we are discussing the analyzed
papers.

Like the analyses, there are few requirements or constraints with respect to a technical presentation. The presentation should last approximately 20 minutes (no longer than 30 minutes). You should be prepared to explain the technical content of the paper you have chosen. Optimally, you would also relate the paper for which you are presenting to other papers assigned for the topic, or more generally to the field. A likely method will be to prepare a PPT presentation in combination with optional use of the whiteboard.

Presentations will primarily be graded on the breadth and depth of the technical content. Secondarily, but still important, papers will be graded based on the quality of the presentation and its effectiveness of communicating the ideas presented in the paper.

For each of the 8 technical topics, I will allow up to two student presentations.

Given that there are more students than available slots, and slots are limited for each topic, assignment of students to slots will be on a first-come, first-served basis. Once the list of technical topics has been announced, you may request one of the two presentation slots without necessarily determining the paper about which you will present. (See below for the process for selecting a paper.)

The paper on which you will present *cannot* be the main
paper for the technical topic (the one about which I am presenting)
but can be another required paper, an optional paper (which will
then become required), or a paper not on the list but that you think
is relevant (subject to my approval--at which point I will add it to
the list).