74

Many of my computational scientist colleagues used to use Google Reader to share and discuss new journal articles. The loss of Reader's social features killed that, and we have subsequently tried Google+ and Reddit, but neither seems to work nearly as well as Reader did for holding this kind of discussion. Does anyone have experience using a site they like for this purpose?

I'm aware of a few options, like http://annotatr.appspot.com/, that seem promising but appear not to be actually used much.

Improve this question
13
  • 1
    There was a site called Phygg that aimed to do this for papers on arXiv, but it shut down due to low participation.
    – David Z
    Commented Feb 1, 2012 at 22:06
  • Can you describe a little better how the discussion went on Reader and what Google+ is lacking?
    – Bill Barth
    Commented Feb 1, 2012 at 22:49
  • 1
    With Google Reader you'd get tightly connected groups wherein one person in the group would share a paper from the journal/arxiv RSS feed and then a number of people would comment on it, often prompted by questions posited by the original sharer. These comments would be semi-private based upon how many people the original poster shared it with.
    – Peter Brune
    Commented Feb 2, 2012 at 2:05
  • 1
    There have been quite a few attempts at providing a comment/review system overlay on top of the arXiv, including scirate.com (defunct) and science-advisor.net.
    – ihuston
    Commented Feb 3, 2012 at 12:03
  • 2
    Why not use blogs? You can make them private if you want only group members to have access.
    – Artem Kaznatcheev
    Commented Jul 10, 2012 at 17:38

8 Answers 8

Reset to default
14

Might I suggest http://scicomp.stackexchange.com ?

While the StackExchange system isn't the best - and indeed isn't designed - for "discussion", I've found that many "What did you think of this paper" type questions can be phrased in SE-compatible formats. CrossValidated has a semi-periodic "Journal Club" bit, and questions and musings about scientific papers come up a fair amount there.

I think if framed correctly, they might find a useful home here.

Generally though, I think the online discussion of scientific papers suffers from a few problems. Generally, the two I find the most problematic:

  1. Lack of a clear community to talk about papers in. Essentially, the problem your question is looking for an answer to. I haven't found a really good general purpose one, though I would love to if I did find it. There's blogs and the like, but even the ones talking about peer-reviewed papers are somewhat one sided in terms of their communication, and not great for anything but transient chatter.
  2. A hesitation to talk about that online. Among colleagues, it seems somewhat easier to summarize things like "Bad paper is bad", or slice apart someone's methodology. I'd be somewhat more hesitant to do that anywhere where my identity is both traceable and the conversation is saved for eternity (the internet).
Improve this answer
6
  • 8
    There is value in being able to be blunt with colleagues about your opinions on papers and techniques. I don't think that SciComp is the right place for that.
    – Jack Poulson
    Commented Feb 2, 2012 at 0:26
  • 5
    I think the lack of anonymity in a public forum will be a deal-breaker for many due to academic politics and fear of reprisal, even if a review is balanced, tactfully pointing out both strengths and weaknesses. People prefer closed fora to avoid this drawback.
    – Geoff Oxberry
    Commented Feb 2, 2012 at 6:13
  • 1
    @JackPoulson I think that it depends on what you mean by 'discussion'. SciComp is probably the right place for certain types of discussion. Things like 'In the paper by EpiGrad et al., can someone explain to me what they mean by X'?
    – Fomite
    Commented Feb 2, 2012 at 7:14
  • 6
    @Epigrad: I think the idea is to do a serious discussion of the intellectual merits and disadvantages of individual papers, as well as areas of research. That's somewhat orthogonal to the goal of SE.
    – aeismail
    Commented Feb 2, 2012 at 9:41
  • 2
    It is my impression that the Scicomp SE is for specific questions pertaining to (mostly) objective technical knowledge and not for subjective value judgments journal papers or research programs.
    – Paul
    Commented Jul 13, 2012 at 14:58
8

I haven't used it personally so I can't vouch for its quality, but I know several people in another research group using a site called Journal Fire for this purpose. Might want to check it out.

Also, I think the citation manager and social network Mendeley has some limited discussion capabilities, but I prefer to manage my references with BibDesk so I haven't use it much.

Improve this answer
3
  • +1 for JournalFire -- it looks good, if I can just get my collaborators to use it too. I use Mendeley, but not for discussions; its interface is not well-suited.
    – David Ketcheson
    Commented Jul 12, 2012 at 6:22
  • 7
    "After five years of connecting researchers at hundreds of universities and research institutes, we regret to announce that JournalFire has shutdown permanently as of October 15th, 2013." - from Journal Fire website
    – mhwombat
    Commented Jul 8, 2014 at 17:38
  • JournalFire is dead now.
    – Scientist
    Commented Jul 11, 2019 at 16:20
6

For public discussions on arXiv preprints there is SciRate.

For general discussions, also private, there is PeerLibrary.

Improve this answer
5

We have just extended BibBase.org to allow papers to be discussed in a fashion heavily inspired by StackExchange: http://bibbase.org/blog/stackoverflow-inspired-scientific-discourse

What's different about BibBase compared to several other sites is that authors integrate bibbase directly into their own homepage. It keeps links to collaborators up to date, and it links to pages on bibbase.org for keywords, and now also for discussion. We want it to be the unobtrusive research network that just helps scientists show their publications online on their own pages as they normally would, but with additional features that make it more than that. We think that it is important to facilitate discussion about one's own papers, and therefore this is now a feature provided by bibbase.

[This is an old question, but it seems that most listed solutions have been shutdown by now one way or another.]

Improve this answer
3

The Selected Papers Network is a new effort in this direction that intends to federate content from all over the web. To post something to it, just write a Google+ post with #spnetwork and the paper's arxiv ID or DOI in the body. You can also post things directly at the site. The developers are working on interfacing with other social tools like Twitter.

You can read more about the thinking behind it here: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3264905/.

Edit: The site was shut down several years ago.

Improve this answer
2
  • 2
    This site isn't working for me. After clicking the link it just redirects to a page that says 'site is starting... click here' in the top left corner. Clicking that link just reloads that page.
    – NauticalMile
    Commented Jan 13, 2017 at 17:00
  • 1
    Google+ being defunct, not sure this is useful now.
    – Scientist
    Commented Jul 11, 2019 at 16:22
2

I use I, Librarian. It is a reference manager, kind of like Mendeley but with a free option.

They have both a paid service option and a free self host option (like wordpress).

It can be private and they have per paper discussion capability.

For a private/semi-private group of collaborators I think it is quite decent.

Improve this answer
0

For theoretical physics (or closely related) papers, one possibility is to use the Reviews section of PhysicsOverflow (note that it is possible for the registered PO users to submit there papers for review).

Improve this answer
0

Wikiversity has useful technical features (the same as Wikipedia): collaborative editing, referencing tools, discussion pages, version control, email alerts, etc. It can very well be used for discussing scientific papers.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .