Wednesday, August 25, 2010

Conference/Journal Versions -- Transactions on Networking

I was recently asked to review a paper for Transactions on Networking, and noticed the following bit in the e-mail?

Please note that while this paper may have had a previous conference version, ToN does not mandate any specific differences between conference papers and their versions subsequently submitted to the journal.

Is this new???  Am I reading this right, that there's no mandated "30% new material", or some similar rule  It's been a while since I've submitted to ToN, but I seem to recall being explicitly asked by reviewers or editors from ToN before what "new material" there was in the paper over the conference version.  I'd be interested to know if this was an actual policy change -- it's one I've called for before, but didn't expect to see implemented anywhere.  

Just curious if anyone can share any insight....


Mike Freedman said...

This change happened maybe 2 years ago, and it's unclear how the policy is realized by reviewers (i.e., are they actually okay without 30% changes)?

This policy seems strongly related to TON's lousy page limit, and it arose around the same time that the "free" page limit was actually shortened: 10 pages for free, then you can buy up to 14 pages.

So, when our SIGCOMM'07 paper was selected for "fast" tracking to TON -- only took 13 months to "review"! -- our final TON paper had to exist in the same amount of space as the SIGCOMM version. We ended up changing about 30%, but, if we hadn't, I'm really uncertain what purpose the journal is serving here. And I've reviewed several submissions from others that were almost *identical* to their conference proceedings. (I don't fully blame the authors here, as they were following policy, and were pushed to redundant journal submissions given tenure requirements at their host institutions.)

In this age of electronic dissemination, the IEEE's hard page limit to keep publishing costs down is backwards looking. They should adopt TOCS's policy and not have *any* page limit.

Anonymous said...

I haven't reviewed for TON but the policy for TOSensorNetworks has some language about "if the editor judges that the revision contains significant amplification or clarification of the original material, or there is a significant additional benefit to be gained from the journal publication."

This is at least better than the 30% new material or other seemingly semantically void criterion (if I change all my notation, does that count? If I include the proofs, does that count?)

Good to see TON is shifting its policy. Perhaps this is because TON gets submissions from very selective CS networking conferences as well as slightly less selective EE networking conferences, and one-size-fits-all was not appropriate?

I think the "no page limit" thing is a mixed bag -- a lot of IT papers are significantly longer than they need to be. It renders the papers almost unreadable in some cases.

Anonymous said...

I believe the reviewers are still considering the 30% rule in their decisions.

I submitted a TON paper in the last year or two that had around 20-30% new material, depending on how you measure new. The reviewers' main criticism was that it was not different enough from the conference version.

Anonymous said...

I agree with Mike Freedman's comment about eliminating page limits in light of electronic dissemination. Publishing an entire conference proceeding (and handing them out to all conferences participants) is so wasteful. At best, I use the proceedings to prop up my monitor. If there is a paper I want to read from a conference, its just easier to find it on the web and (maybe) print it.

In writing this, I realize there are many aspects of the publishing business that I do not understand, so I plead ignorance in knowing the reasons for hard-copies of conference proceedings. Perhaps, someone can explain. Nonetheless, complaining is therapeutic, so thank you for allowing me to indulge.

Anonymous said...

A related question is the following: what if you get a ToN submission with an identical tech report that has been public for a long time (e.g., 2-3yrs)? Would you reject saying it is "identical"? Would TRs count as publications in this regard?

Mike Freedman said...

@ Anonymous (August 27, 2010 10:05 AM)

USENIX has started electronic-only publishing even at its top venues, including NSDI '10 and the forthcoming OSDI '10, forgoing the normal printed conference proceedings.

At NSDI, at least, they distributed a memory stick at the conference with all the papers preloaded, and papers were immediately available to attendees online.

On the side note: USENIX is a great organization if you aren't familiar with it, with one of the more enlightened copyright policies (compared to the ACM and the IEEE).