Ladies and Gentleman,

My class presentation today is based on the following paper:
Defense of Trust Management Vulnerabilities in Distributed Networks Yan (Lindsay) Sun , Zhu Han and K. J. Ray Liu

You can download it from the following link:

www.cspl.umd.edu/sig/publications/sun_CommMagazine_draft.pdf

If at all possible please familiarize yourself with the paper prior to our class today.

Sincere regards,

Oleg Aulov

summaries of recent class sessions

3/5 started 10 minute talks by students

3/3 handed back some homework, and went through slides about how to do presentations

2/27 returned the plagarism exercises, which everybody did well on, for the most part

2/25 If I remember correctly, I talked about resumes and what you should or should not include. That led to topics like discrimination based on gender, national origin, and so forth.

ADOBE AIR

this is a wrapper on basecamp (www.basecamphq.com) that i built when i worked at a software house about 2 months back.

DEMO VIDEO of the AIR APPLICATION:

http://www.youtube.com/watch?v=itb3XpdhrYY

My new blog !

Hey ,
I've created my new/first blog.
If you are interested in blogging just about anything, do visit my blog.
cheers,
GP.

Literature Surveys

Sitting in a lab meeting the other day someone said "literature surveys are kind of a black art", and proceeded to discuss sort of how to go about doing one, from taking an interesting paper, following its references backwards to a seminal paper and then following citations forward from the seminal paper to current work.

I have yet to start a survey for class, but I do have one or two papers that I think are interesting, topic-wise. Does anyone have any tips or processes they use for doing literature searches/surveys?

Wes

outline of a thesis proposal

1. problem statement, why do you care?

2. what have others done? is there related work?
what are you going to do to go beyond the other work?

3. how would YOU attack the problem? what makes this approach promising?

4. evidence? how would you GET the evidence?

5. research plan - what needs to be done?

Isn't looking for an opening first is better than looking for an idea

how do we find out what people want ? , example what a company is looking for , which field are they willing to invest , which algorithm are they most interested in improving etc. ?

It would be much more rewarding to start from there and then try to solve a problem rather than look for a scattered idea in a field.....

The Academic Career

On Wednesday we started discussing details of the academic career, such as salary expectations, appointment coverage, and negotiation during hiring. I've been doing some personal research on these topics, mostly because I'm using my degree to switch careers from industry into academia and I'd like to know up-front what the academic career is like, especially compared to industry. I honestly was surprised to learn that a starting salary for an assistant professor is in the low eighties, I was expecting it to be lower than that. I was also surprised to learn that summer support can be negotiated for the first couple of years. I'm hoping we can have some more discussion about this topic. I've been poking around Chronicle Careers and have found some interesting articles about this topic, a recent one being a new column about three new professors.

points to ponder from Zobel, chapter 10

As I was reading Zobel's chapter 10, I was struck by several points. I'm not trying to summarize his chapter, but rather to list some talking points:

there is indeed a process of research, and papers should? and often do have a life cycle

insight is necessary, but only 1% - as Edison is famous for saying

you get to choose whether an insight is worth any investment of time and energy

a clear problem statement is critical - to help keep focus, and avoid distraction

when choosing a project, play to your own strengths (Zobel, p. 161)

every student-advisor relationship is different! students working on their own have their own advantages and disadvantages

we've discussed the importance of reading papers, and you'll get more practice

top of p. 167 has a nice list of questions that may be useful in reading a paper and preparing a summary!

setting milestones is very helpful! Zobel may say necessary!

explain the role of the hypothesis in scientific research - related to problem statement, next level of detail, and they drive the research activity e.g. experiments to run, prototypes to build, surveys to conduct

what are the obvious arguments pro or con with respect to a hypothesis? If the answer is obvious, it may still be wrong! But this may guide experimental design

Zobel's comments about formal studies and well-designed experiments are well-taken, but may be arguable. If a claim is too good to be true...

Checklist on pp. 182-3 may be useful!

1st research paper analysis

on page 407 . right under the 'Average Degree of Seperation' heading the author writes..''i calculated exhaustively the minimum distance..... for all pairs of scientists in our databases for whom a connection exists "

which sounds quiet unbelieveable ... because looking at table
1 .total authors for MEDLINE = 1520251
hence he calculated 1520251 C 2 ..(where c means combination)which = 1.15 * 10 ^ 12
for just one database ...

and then he applied some minimum distance algorithm which has queit a big complexity of its own since the database is so large for EACH AND EVERY PAIR ?
and later he says that he paired all the authors from the databases.... meaning adding all the authors and calculating shotest distance for all pairs ? :S ?

it sounds a little vivid to me because an exhaustive computation of that is very very big, or did I not understand what he said.Can sme1 please help me out.

reflections on Wed. 1/30

Talking about LaTeX seems smooth enough. I should add a pointer to the LaTeX reference card, and give people some exercise. Maybe a resume?

Spent more time talking about descriptive markup than I expected, but that's fine.

Finished Kung, realizing that the format of that isn't quite right. That would go better as actual slides, rather than a long web page to be scrolled through.

reflections on the first day, 1/28/08

It looks like it will be a good class.

Kung's overview, although dated from a technical point of view, is a better topic to start with than plagarism! Which is still an important topic.

Spent more time than I expected talking about different publication venues, and how in some fields conferences are more important than journals and v.v.

Didn't get a chance to talk about LaTeX at all, but that too can keep until Wednesday.

Passed around copies of Newman, and told the class to have it read by Monday of next week - I should have said Wednesday, but I'll have a chance to talk about paper summarization anyway.

Speaking of Wednesday, I want to talk some more about Kung's overview, and I'll talk about LaTeX. Then paper summarization, and that will lead to rules regarding source citation and plagarism.

So, what are the goals and objectives of this course?

Goals/Objectives:

• Take some of the mystery out of the M.S. and especially Ph.D. process - when people understand the process, they're more comfortable being part of it.
• Give practical advice about how to succeed
• Develop basic skills - literature search, problem definition, research methodologies, written and oral presentation, formulating a research proposal

things added since last time

I taught this course a three years ago! That was Spring 2005. Marie taught it again in 2006, and Krishna in 2007. What did they add? What should I add?

Marie added a mailing list, and a wiki. I've set up a blog.

Marie used some AI papers, as examples for summaries. I might use some different papers.

http://www.sce.carleton.ca/faculty/chinneck/thesis.html

http://www.cs.umbc.edu/%7Emariedj/papers/advice-summary.html

I like the CV material, and students need to learn about professional networking. So I'll keep the paper by Susan Eggers, Networking and Professional Social Interaction.

Faculty research presentations are a good idea!

The statistical methods should be good. You might argue that there should be more than a week. One Matlab class is fine, but we could do SAS or some other stat package. TBD.

Tim Finin talked about setting up a web presence, and he had other comments in a ppt.

Looking at Krishna's schedule from last year: pointers to IEEE and ACM Computing Surveys.

I want to add links to the ACM DL, citeseer, Google Scholar, Usenix, HBR, maybe others...

Writing Technical Articles (with emphasis on paper in systems and networks), by Henning Schulzrinne.

Jack Lynch's Guide to Grammar and Style (Last Updated: Dec 2006 as of March 2007)

There's a lot out there on writing technical papers.

Also, lots of LaTeX stuff.

http://www.cs.virginia.edu/helpnet/Being_Grad_Stud/grad_school_CS.html

Welcome to the Spring 2008 CS 601 blog!

Welcome to the blog for the Spring 2008 offering of CMSC 601!

If you are enrolled in this class, you are welcome to post. If you have taught the course before, you are also welcome to post! You may need to register at blogger.com, but that's easy and free.

The purpose of having a blog is to document the course, in terms of topics that may be discussed, special events like snow days, or comments about how well a particular class session went.

This is one of my favorite courses to teach! And we are all lucky that Marie and Krishna have invested so much in getting the course started. The class looks like it may be a little smaller than in previous years, and that gives us some flexibility in terms of maybe adding new topics this semester.