A New Model for Scientific Publication

For the recent AAAI Symposium on Computational Approaches to Analysing Weblogs, I prepared a paper which looked at various findings from 24 hours of posts in the blogosphere. While I did the research for this paper, I posted some findings on this blog. Each post looked at a single experiment and gave summary results and insights. The paper was, to a large extent, a compilation of these posts.

From this, I think there is an opportunity for social media, and blogs in particular, to change in a subtle manner some aspects of how scientific research is disseminated. Here is the model (version 0.0):

1. An author posts a small piece of research on a blog.
2. The post is reviewed by some number of peers either via comments or some other mechanism.
3. The post is archived by an aggregator in a manner which makes the post persistent, searchable, etc.
4. The post is cited in other posts, in formally published articles, etc. using a standard format for referencing articles published in this manner.

This proposal would probably need something like a microformat to support the aggregation and possibly the citation process.

A key to getting this to work is to figure out step 2. - how would peers be selected to review and how would the reviews become part of the aggregated meta data? I suspect that there are some interesting social network effects that could facilitate this based on self selection. It is possible that such a scheme could be built on top of a paper aggregation and search system like Rexa.

Comments

[paolo]

1) what about people arguing "someone is going to steal my idea if I published it on the blog and it is not yet published on a 'real' paper"?
2) Check http://hublog.hubmed.org/archives/001362.html
3) And I guess you know already citeulike (which I prefer) http://www.citeulike.org/ and connotea (which is Free Software) http://www.connotea.org/

Ideas? ;-)

[Matthew Hurst]

Paolo,

Thanks for the comments.

1) this proposal would certainly present results in something equivalent to a 'real' paper. I.e. publishing results in this type of medium would be an acceptable thing to do (which is why getting the referencing system right is important).

2) Looks interesting - One thing I hadn't considered was the bibliography, it would have to be part of the microformat.

3) I was aware of these, but they provide something a little different to what I propose.