## Thursday, August 03, 2006

### Musings on the arXiv.

Things that make you go 'hmmmm', or why arXiv can't replace peer review:
The polynomial solution of graph isomorphism problem is obtained by consideration symmetry properties of regular $k$-partitions that, on one hand, generalize automorphic $k$-partitions (=systems of $k$-orbits of permutation groups), and, on other hand, schemes of relations (strongly regular 2-partitions or regular 3-partitions), that are a subject of the algebraic combinatorics.
Although I appreciate the value of the arXiv, there are some things about it that I still don't get.
Take the paper excerpted above. It claims to solve graph isomorphism in polynomial time, a result that if true would be quite momentous. The abstract already warns me that the technical material would be complex, and (call me parochial, shallow, what-have-you) the complete absence of grammatical structure does raise some tiny red flags.

So where does this leave me ? I'd like to know whether there is any merit to this paper. In the absence of peer review, I'd have to wade through it myself, or hope that someone with more technical expertise in the material does it for me, and announces their verdict in a forum where I can hear about it.

Categories:

1. ...hope that someone with more technical expertise in the material does it for me, and announces their verdict in a forum where I can hear about it.

If this paper appeared in ECCC, which allows posting comments, your wish would have been granted in a blink of an eye.

Piotr

Posted by Anonymous

2. So the arxiv *does* allow a weird kind of commenting. If you reference a paper in your blog, and set up trackbacks appropriately, the note will appear alongside the paper. I know this works for the physics arxiv, (and is somewhat controversial because they "censor" comments), but I don't know if it was folded in for CS.

Posted by Suresh

3. This is the one that (as I mentioned briefly a couple days ago) towards the end says something about how one should read the other arxiv papers by the author to understand more of the development of the theory, but that they are "not mistake free". Which was enough to convince me that I should hold off reading until the author could no longer be so confident about the existence of mistakes.

If you think of arxiv publication as similar to tech report publication you won't be far wrong I think: a way to conveniently collect papers but until they list a journal or conference reference there's no guarantee of any kind of peer review beyond a very basic level of checking for whether it's on-topic by the arxiv mods.

Posted by Anonymous

4. ECCC has a screening process whereby a paper can be posted only if one of the board members "accepts" it, as opposed to the arxiv model where a moderator can remove an off-topic paper after it appears. The screening is not meant to verify correctness, but a paper with extravagant claims would probably receive careful scrutiny from an expert. Incidentally, ECCC, unlike the arxiv (which has both types), has no article proving P=NP and no article proving P!=NP.

Posted by Luca

5. A good (though rare) example of an ECCC discussion on a problematic paper is TR06-062. (I believe the discussion is not finished yet - anything to add?..)

Posted by edwardahirsch

6. I see the purpose of submitting a paper in ECCC is a way to receive peer-review comments. I am not sure if it is exactly inverted way of using it. For me, neither do I see it embarrassing nor do I intend to submit so that I can claim that the "proof belongs to me"(hard to prove this claim though :) , but following explanation might serve).

I do not belong to academia. And I really do not get much opportunity to get my works discussed with others or reviewed. While I do not indent to live with that. Thus I rather decided to post in ECCC – as even one constructive comment will be worth for me, as that will allow me to correct my work quickly. It is rather in this desperation the problematic paper’’ exists as TR06-062. While I must say I am thankful to, Chris Calabro and Yann Barsamian, who commented on the earlier version, and I am working on them. Few other things that I would like to point out are following:
1. There were published (conf.) papers where claims were wrong.
2. It is nearly impossible to receive any useful comments on a completely wrong paper.
3. While part of a work might be wrong – but some ideas presented might be useful, and I think that is the basic idea of ECCC like archives.

Posted by Subhas