Distributed or How to escape from Grad School

2539 days ago, 825 views
PowerPoint PPT Presentation
Why distribute?. To go to colorful hotelsTo inspire your mom with your name in printTo graduateexternal reviewTo get a jobyour consultant thinks every one of his understudies are above averageTo fulfill exploration contract necessities. What number of papers do I have to graduate?. 1 Science paper or

Presentation Transcript

Slide 1

Distributing – or How to escape Grad School Henning Schulzrinne Dept. of Computer Science Columbia University (refreshed September 2009)

Slide 2

Why distribute? To go to extraordinary lodgings To awe your mom with your name in print To graduate outside audit To land a position your counsel supposes every one of his understudies are better than expected To fulfill examine contract prerequisites

Slide 3

what number papers do I have to graduate? 1 Science paper or … 2 Sigcomm papers or … ~5 "genuine" productions

Slide 4

Publications Different sorts of distributions Think like an analyst Finding the correct meeting Advertising your work Paper sorts What if my paper is rejected?

Slide 5

Publication sorts Technical reports, including arXiv Workshops Conferences Magazines ("Archival") Journals Internet Drafts and RFCs

Slide 6

Finding out about meetings CFP = get for papers Finding out about gatherings http://www.cs.ucsb.edu/~almeroth/conf/details/http://wikicfp.com TCCC declaration list (subscribe!) SIGCOMM and ComSoc website pages

Slide 7

Technical Reports CS, IBM and BL TR arXiv.org abstains from being "scooped" introduce extra points of interest (reproduction comes about, verifications, usage subtle elements) Can be utilized to promote on mailing records – read more frequently than some meeting papers

Slide 8

Workshops Two sorts: welcomed (Dagstuhl) subject centered ("Internet Measurements", NANOG) Smaller, more engaged than gatherings May not have formal procedures, just duplicates of slides Often, just on more than one occasion, yet some for quite a long time (ComSoc) Selectivity differs – from 100% to 10% Some occasions are called workshop, however are truly gatherings (NOSSDAV, IWQoS)

Slide 9

Conferences Hundreds a year Traditional: ICC, Globecom Semi-customary: Infocom, SIGCOMM, ICNP, Sigmetrics, Usenix, SOSP, … Newer: WWW, NOSSDAV, IWQoS, SAINT, Mobicom, Mobihoc, … Submission estimate: 5-12 pages

Slide 10

Conferences Some have short entries ("expanded unique") and longer acknowledged papers Some are successfully a similar length (Infocom) Few have long entries and shorter last papers

Slide 11

Conference surveys Either specialized program panel (TPC) or TPC + outside commentators Reviews dazzle (most IEEE gatherings): writer doesn't know analyst, yet analyst knows writer personality twofold visually impaired (ACM): just the seat knows the writer characters

Slide 12

Finding the correct meeting Filter #1: just submit to meetings filed in IEEE eXplore or ACM DL aside from possibly for a few workshops Appropriate gathering layer/theme territory style (examination, framework) selectivity area (Australia versus NY)

Slide 13

Traveling to gatherings Many bigger meetings have understudy travel gives frequently for writers here and there for non-writers (SIGCOMM)

Slide 14

Magazines Examples: IEEE Network Magazine IEEE Communications Magazine IEEE Wireless Communications IEEE Multimedia Magazine Large flow  subjects of expansive intrigue Written for non-master (30,000 perusers!) Originality not generally most essential

Slide 15

Journals Every PhD proposal ought to bring about no less than one diary production Archival – most libraries have them and keep them everlastingly Long audit cycle Selectivity differs enormously – can be less specific than a few meetings Often, given additional opportunity – "resubmit with significant changes"

Slide 16

Journals Issued mainly by Societies ACM IEEE Commercial distributers Springer Verlag Kluwer North Holland

Slide 17

Journals Examples IEEE/ACM Transactions on Networking Journal on Selected Areas in Communications Computer Communications Review (CCR) ACM Transactions on Multimedia Computing, Communications and Applications Computer Networks Journal of High Speed Networks Journal of Communications and Networks …

Slide 18

RFCs – Internet Standards Documents RFCs are not papers (and the other way around) Can take a while, especially for measures track reports Start with submitting Internet Drafts – yet most Internet drafts never make it to RFC Specification versus depiction

Slide 19

RFCs Precision versus oddity and execution "How can it work" versus "how is this superior to existing work" Good approach to get affect Good for mechanical prospective employee meet-ups

Slide 20

Ways to publicize your work Technical reports Put connection and theoretical on website page (web crawlers!) Relevant mailing records (e.g., end2end) Send pointer to creators of work that is firmly related arXiv for tech reports

Slide 21

Finding related work netbib citeseer Google site pages of surely understood system look into gatherings Digital Library, IEEEXplore

Slide 22

Types of papers - content Measurement measure execution of genuine frameworks test bed or genuine Internet watchful insights – how illustrative is your information? Investigation of existing calculation TCP, FDDI, DQDB, RED, … - not some dark convention recreation or examination awful conventions are uplifting news for creators…

Slide 23

Types of papers, cont'd. Framework depiction execute intriguing framework portray it in adequate detail what's new and fascinating? model, not mechanical item New calculation or convention exchanging, steering, planning, … execution assessment most elevated hazard/compensate don't depict bit fields

Slide 24

Think like an analyst Reviewers are volunteers Reviewers are not English editors conclusion: "in the event that you can't utilize a spell checker, why should I put stock in your diagrams and conditions?" Abstract and title need to guarantee appropriate directing of paper (hypothesis versus frameworks) don't overpromise: "take care of QoS issue" versus "add change to DiffServ to better serve soccer recordings" Reviewers get frantic if their work is not refered to Clearly state what your commitment is (and state different things in future work)

Slide 25

Think like an analyst, cont'd. Clear inspiration – essential for non-master commentators "is the issue vital?" Sufficient detail to assess, however not "utilized gcc 1.2.3 on a SPARC Ultra 10 called snoopy to recreate" Avoid nonexclusive inspirations "The fast advances in foo"  antique! Rehash fundamental outcomes in presentation and rundown end product: papers are not tension books – should have the capacity to see degree, inspiration and results on first page Very deliberately recognize from earlier work including your own earlier work ! Abstain from over-burdening one paper (hard!) paper ought to recount a story, not be an examination index or cerebrum dump

Slide 26

Paper accommodation Technical report (and RFCs) do no mischief Basic run: can't submit same material to two scenes all the while ( including gathering and diary) = twofold accommodation Don't investigate LPU Conference paper = refined(workshop paper + detail) Journal paper = refined(  meeting papers)

Slide 27

What if a paper is rejected? Try not to hop off the GWB - it happens to everybody If not, you're not submitting to the correct gatherings No point griping if the audits are shallow – choices are viably last (aside from revelations of written falsification, and so forth.) Publish as tech report promptly (in the wake of thinking about surveys)