I completely agree with your idea of engaging reviewers in a dialog. There’s a lot of mis-understandings that happen with blind reviews and I would often have preferred transparency both as a submitter and as a reviewer.
But I’m a little worried by your proposal – if it takes off. Suppose everyone did this, including Peter Norvig. Now, say I have this cool idea for fast integer decoding – who am I going to send it first: Norvig or Lemire? A simple game theoretical analysis says Norvig, even though Lemire is clearly the better choice. Why? Because Norvig might get me a high-paying job with a cool company and has thousands more Twitter followers than Lemire.
I think that keeping the function that journal editors play as “switchboards” for peer-matchups might still be better than “send me your draft, and I’ll advertise your paper if you revise it to my satisfaction.” We just need to change how journals work.
I see a business opportunity: an “E-harmony” for matching authors and reviewers. A technology development opportunity for a better kind of recommender too – reviewers could be “rated” and recommended across different dimensions (subject matter knowledge / detail of reviews / helpfulness). Authors could be rated according to how finished their submissions are (I hate it when I get a poorly written submission with mistakes everywhere in order to get free proof-reading)….
In the future I imagine, you’d seek a reviewer for your latest manuscript, maybe using some kind of computer-assisted process. You’d obviously want your reviewers to be as well known and as influential as possible. But this is actually a good thing!
I don’t imagine this would replace journals. It is more about going beyond journals.
That’s like saying “YouTube won’t replace TV channels”. No, it won’t.
By the way, though I can only guess, I suspect that Norvig might be open to something like I describe. Google has not focused its research on journals and conferences.
In my opinion we would (at the very least) need to (1) differentiate (i.e. givign different weigths) between pseudonym and “real name” users and (2) differentiate between pure up/downvote and actual reviews/comments.
Yes I am very much interested in this.
Open reviews. Reviews of reviewers. Some closed reviews if really needed. Repositories/versioning for papers.
At the end only citations really count. Reviewers can help a paper get better IF there is no competition. Ideally… but we all know this is not true.
So better leave it to the judgement of the masses. Crowdsource reviews and comments.
At the end only the number of citations and impact of the paper really count.
I completely agree with your idea of engaging reviewers in a dialog. There’s a lot of mis-understandings that happen with blind reviews and I would often have preferred transparency both as a submitter and as a reviewer.
But I’m a little worried by your proposal – if it takes off. Suppose everyone did this, including Peter Norvig. Now, say I have this cool idea for fast integer decoding – who am I going to send it first: Norvig or Lemire? A simple game theoretical analysis says Norvig, even though Lemire is clearly the better choice. Why? Because Norvig might get me a high-paying job with a cool company and has thousands more Twitter followers than Lemire.
I think that keeping the function that journal editors play as “switchboards” for peer-matchups might still be better than “send me your draft, and I’ll advertise your paper if you revise it to my satisfaction.” We just need to change how journals work.
I see a business opportunity: an “E-harmony” for matching authors and reviewers. A technology development opportunity for a better kind of recommender too – reviewers could be “rated” and recommended across different dimensions (subject matter knowledge / detail of reviews / helpfulness). Authors could be rated according to how finished their submissions are (I hate it when I get a poorly written submission with mistakes everywhere in order to get free proof-reading)….
@Andre
In the future I imagine, you’d seek a reviewer for your latest manuscript, maybe using some kind of computer-assisted process. You’d obviously want your reviewers to be as well known and as influential as possible. But this is actually a good thing!
I don’t imagine this would replace journals. It is more about going beyond journals.
That’s like saying “YouTube won’t replace TV channels”. No, it won’t.
@Andre
By the way, though I can only guess, I suspect that Norvig might be open to something like I describe. Google has not focused its research on journals and conferences.
Could it be built upon http://arxaliv.org/ ?
It’s already there and it’s open source.
In my opinion we would (at the very least) need to (1) differentiate (i.e. givign different weigths) between pseudonym and “real name” users and (2) differentiate between pure up/downvote and actual reviews/comments.
Yes I am very much interested in this.
Open reviews. Reviews of reviewers. Some closed reviews if really needed. Repositories/versioning for papers.
At the end only citations really count. Reviewers can help a paper get better IF there is no competition. Ideally… but we all know this is not true.
So better leave it to the judgement of the masses. Crowdsource reviews and comments.
At the end only the number of citations and impact of the paper really count.
I wrote about this also here:
https://engineering.purdue.edu/elab/blog/publications/publishing-model/
You might also be interested in:
http://openreview.net/
Some of this is coming!!!