Wednesday, October 29, 2008

hack to break reviewer anonymity

a rather straightforward hack to break the anonymity of reviewers for peer-reviewed publications (most places at least ensure single-blind submission -- author names are public but reviewing is anonymous)


wonder if this has been tried before to figure out who is reviewing your paper?
1. put some results/theorems in a separate file (preferably these are genuine)

2. give a url submissionid_conference.googlepages.com/filename for the file (probably works much better if the submission is double blind -- if author names are also suppressed and all hints to author identity are explicitly removed, then you can just create something and justify why you created a webpage for the additional stuff from 1 -- linking to your homepage would sacrifice author anonymity :-) )

3. add some webtracking thing to the url and see where you are getting hits from

assuming you can filter out spurious hits/click spam and check for actual views (e.g., .edu or city names), you should have a fairly reasonable idea of who is reviewing

(inspired in part by a recent paper that cited itself for motivation :-) )

3 comments:

Harsha V. Madhyastha said...

"cited itself"?!! Or, did you mean cited a previous paper by the same authors?

nice try said...

no i meant "cited itself"

the submission version of a paper cited a tech-report version of the submission in the intro as part of problem motivation

Samyukt said...

what do u do when someone snoops into the laptop and unearths all secrets ???!!!........nicertry blog, Japan visits etc were serendipitous byproducts of unabashed data mining done with utmost finnesse !!!!!!!!!!