Dear all,

Here a proposal that aims at reducing review workload. The idea is to split the review of a paper in two stages. It implies that each paper has two parts:
- an abstract aimed at the non-specialized reader clearly stating the contribution of the paper (selling it, actually). So the abstract should probably be longer that what we have now (say a 2-page limit)
- a technical part that will typically be more specialized

Stage 1: the paper is reviewed by a relatively large number of people from different sub-disciplines, based on the abstract only. Reviewers should assume that the technical part will deliver what the abstract announces. If the paper survives this phase, it proceeds to phase 2.

Stage 2: a few specialized reviewers check in detail whether the technical part delivers on the promise made in the abstract. If so, the paper is accepted. This includes verification of proofs, claimed attack complexity etc.

At least in theory this may reduce the workload as most reviewers only have to read the abstract. Forcing the authors to write an abstract aimed at a wider audience has the additional benefit that papers may become more accessible to people working in other sub-disciplines.

I realise that whether this really works depends on how it is implemented. For example, something must be built in against overselling. This could be done by having a system with (negative) points where each co-author gets a point when his paper passes stage 1 but not stage 2 and these points are somehow taken into account in stage 1. And of course there are many other details that may make this a success or a failure. But let's first see if there is support for the basic idea in the first place.

Joan

But in this case, if you accept X papers, you need to pass to Stage 2 X+Y papers (assuming some will fail the testing), and then, you are susceptible to variance (Y should be O(X), IMHO) in the number of accepted papers, or some decent papers gets rejected due to lack of space.

In addition, why not to just separate the submission into two parts: abstract, and the rest. Then, interested committee members could immediately check the details if they wish. Which is btw, what happens now, but with shorter abstract.

We don't have space limitations anymore - we can accept any good paper from now on :-)

Best,
Christopher

Orr Wrote:
> you are susceptible
> to variance (Y should be O(X), IMHO) in the number
> of accepted papers, or some decent papers gets
> rejected due to lack of space.

We have space limitations in the conference itself...

True. But it's about the IACR Proceedings. Here, we do not have any space limitations. Just quality restrictions.

For the conferences, you are 100% right though.
Christopher