Answers for Assignment | Software Engineering | CS 420, Assignments of Software Engineering

Material Type: Assignment; Class: Software Engineering; Subject: Computer Science; University: University of Alabama - Birmingham; Term: Fall 2005;

Typology: Assignments

Pre 2010

Uploaded on 04/12/2010

koofers-user-sdl
koofers-user-sdl 🇺🇸

9 documents

1 / 2

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
1) We find the "balance" policy vague and confusing.
Answer:
Balance here means that:
One precondition to all following procedures: a reviewer cannot review his own paper,
even though he is not the first author of the paper.
a) Assign papers to the reviewers whose major research area matches papers' keyword(s).
(See the example, every reviewer has only one major research area, and papers can have
one or more keywords).
b) Check the balance of all reviewers
c) If not balance, the assignment system should find the reviewers who have more
workload. Assign some of these reviewers' papers to other reviewers whose minor
research areas match the papers' keyword(s) to achieve the balance policy.
The definition of balance is that the number of papers each reviewer review should not
exceed 2 or more papers than the average (average=papers/reviewers).
d) Check the updated balance of all reviewers
e) If still not balance, the program committee chair will adjust the workload manually. --
PC chair has no limitation for assignment except the precondition!!
2) For keywords for submission, are these provided or can they type in anything?
Answer:
If you have a GUI for submission, authors have to type in after proper login. If you want
to use a DATA file for each author to input all information for paper submissions, it can
be predefined. I think both approaches work. But I am worried about the demo time issue
of the GUI approach. You may try both approaches during demo.
3) When are papers assigned?
i. as they are submitted
ii. after a percentage of total expected papers are in
iii. after a pre-chosen number are in
iv. after a pre-chosen date set before the deadline
v. after the submission deadline
Answer:
All papers are assigned after the hard deadline pre-defined by the PC chair. (the answer is
v)
4) How are keywords determined?
i. author selects all that apply from a list
ii. author types keywords into a comment-style box
iii. are keywords nothing more than selecting from the Areas of Research provided.
Answer:
Keywords are determined by the authors. To make it simple, all keywords selected have
to be in the list (8 keywords) stated in the first paragraph of page 1. (the answer is i).
5) For paper submission, no paper can be submitted after the submission deadline. This
means that the deadline is the key point, to split the submission and assignment activities.
pf2

Partial preview of the text

Download Answers for Assignment | Software Engineering | CS 420 and more Assignments Software Engineering in PDF only on Docsity!

  1. We find the "balance" policy vague and confusing. Answer: Balance here means that: One precondition to all following procedures: a reviewer cannot review his own paper, even though he is not the first author of the paper. a) Assign papers to the reviewers whose major research area matches papers' keyword(s). (See the example, every reviewer has only one major research area, and papers can have one or more keywords). b) Check the balance of all reviewers c) If not balance, the assignment system should find the reviewers who have more workload. Assign some of these reviewers' papers to other reviewers whose minor research areas match the papers' keyword(s) to achieve the balance policy. The definition of balance is that the number of papers each reviewer review should not exceed 2 or more papers than the average (average=papers/reviewers). d) Check the updated balance of all reviewers e) If still not balance, the program committee chair will adjust the workload manually. -- PC chair has no limitation for assignment except the precondition!!

  2. For keywords for submission, are these provided or can they type in anything? Answer: If you have a GUI for submission, authors have to type in after proper login. If you want to use a DATA file for each author to input all information for paper submissions, it can be predefined. I think both approaches work. But I am worried about the demo time issue of the GUI approach. You may try both approaches during demo.

  3. When are papers assigned? i. as they are submitted ii. after a percentage of total expected papers are in iii. after a pre-chosen number are in iv. after a pre-chosen date set before the deadline v. after the submission deadline Answer: All papers are assigned after the hard deadline pre-defined by the PC chair. (the answer is v)

  4. How are keywords determined? i. author selects all that apply from a list ii. author types keywords into a comment-style box iii. are keywords nothing more than selecting from the Areas of Research provided. Answer: Keywords are determined by the authors. To make it simple, all keywords selected have to be in the list (8 keywords) stated in the first paragraph of page 1. (the answer is i).

  5. For paper submission, no paper can be submitted after the submission deadline. This means that the deadline is the key point, to split the submission and assignment activities.

  1. Can a paper be reviewed by a reviewer more than once (for instance, if they want to change something)? Answer: No in both real and virtual worlds.

  2. Three extreme cases: a) It says that the at least one paper is accepted from each Research Area. What if none are submitted? b) If any are submitted, then one will get the high score (even if it is a score of 0), right? If ten have a score of zero and the acceptance policy is 50%, the 1st five zeros win. c) If none are submitted, I vote for the Blue Screen of Death! Answer: These three examples are extreme cases rare happen. For the first one, I will say the requirement of all areas have to have at least one paper can be ignored. For the second one, that is true. But PC chair can use his judgment. For the last one, we have nothing to test in this case, you may output a warning or something.

  3. If a paper isn't marked as acceptance does that mean it is automatically rejected? Answer: That is right. This will be a boolean variable

  4. Some of us think it is a piece of software that always runs and does things automatically at appropriate times. Others of us think of it as the entire package. For instance, how does evaluation get started. It is not stated in the requirements, but I think the PC Chair should use an interface tool to start this process. Otherwise, the "Virtual PCC" must automatically do this.. Answer: I don't think the VCC is an entire automatic system. Some interactions between authors and VCC and PC chair and VCC exist. It is appropriate for PC chair to invoke the assignment, evaluation, scheduling, and notification processes. Also PC chair has right to modify any result of these processes after being decided by software - a simple interface for updating values will be fine.

I forgot to mention that there should be a deadline for splitting reviewing and evaluation activities as well. After all reviews are reviewed by the deadline, the PC chair will request the system to evaluate the scores and decide the final accepted papers. Of course, before scheduling system starts, PC chair can manually change the results. Then the PC chair requests scheduling.