Limitations: (1) 140 chars is too limited to accomodate a typical URI, (2) needs to be coupled to author ID, (3) only one rating dimension as of now. - Daniel Mietchen

if we could aggregate a whole bunch of different comments from different sources and translate them automatically to fit? Or provide a link to the comment rather than the full text of the comment? - Cameron Neylon from twhirl

Ok, now I've looked properly. Would probably need to build a slightly more intelligent service but if you could parse links then there are two very nice things here. One is linking id, and review via URLs but the other is that this actually matches the PubSub-Syndicate mechanism that Jon Udell talk a lot about much better than commenting on websites (...now where is that half written blog post...?) - Cameron Neylon

the comment text from people usually contains even more helpful information than rating metrics - Mike Chelen

Daniel: what about linking by DOI? - Mike Chelen

DOIs can still be very long - plus you'd probably want to give them in the http form so that makes them even longer. Mike, agree the text can be more useful but if it is held somewhere else then the tweet only needs to reference it via a shortened link was what I was thinking - Cameron Neylon

The DOI problem could be solved if the @hreview service were to expand the shortened URL per default. - Daniel Mietchen

Also, we need a URI scheme for anything on the web, from blog entries to wiki edits to @hreview ratings. - Daniel Mietchen

Cameron:probably the posts could contain shortened DOI URLs that were expanded in the underlying hReview - Mike Chelen

Daniel: can the hCard support be used to integrate with other author ID systems? - Mike Chelen

Yes, hCard can do this: http://microformats.org/wiki... . What we need is a functional author ID scheme. - Daniel Mietchen

What would be a suitable license for ratings? http://www.opencritics.de/ use http://creativecommons.org/license... - Daniel Mietchen

CC-BY-ND certainly makes sense for the pure values (text or numbers) of the ratings, but isn't it too restrictive for reuse, e.g. aggregation? - Daniel Mietchen

Daniel: there are clauses to specifically allow collections, but it's never been clear to me exactly how ND applies to subsets or programmatic reuse. - Mike Chelen

Nor to me, Mike, but the opencritics are open to criticism (I had to bring that) and suggestions, and have lawyers to sort such things out properly. More to come on that by tomorrow - just had them on the phone. - Daniel Mietchen

Examples of CC-licensed ratings provided by OpenCritics: http://bit.ly/JwQiH (simple), http://bit.ly/42H3l3 (more detail) - scroll down in both cases. - Daniel Mietchen

Further thoughts on such rating schemes for science: http://bit.ly/2xJpKD , http://bit.ly/18jc9o and http://bit.ly/5HzJR . - Daniel Mietchen

Emerging impact measures based on @tweprints stats are described at http://bit.ly/FZp7a . - Daniel Mietchen

hmm, 140 characters won't work. We need an article social activity aggregation service, could then be mirrored onto twitter or where ever. Like stramosphere, but just for articles. Euan should write it. - Ian Mulvany

Should note that CrossRef is thinking of creating alternative shortened DOIs that could address some of these problems. Working name is "toydoi". Advantage of CrossRef is we could avoid spam-plague faced by traditional URL shorteners. Would be good to hear from interested parties to understand use cases. - Geoffrey Bilder

Ian, I think that "article social activity aggregation service" is a good description for what Mendeley are up to - still a bit rough a toy, but improving very fast. - Daniel Mietchen

Geoffrey, good to read that. However, the length of DOI is just one problem, and more pressing from my point is to develop a DOI-like URI scheme for anything cited in a scientific context (and for anyone citing, too), e.g. via automatic deposit at places like Webcitation or Portico (and using some sort of author ID). And before going public with that working name, they might wish to invite comments from speakers of Vietnamese. - Daniel Mietchen

So we've also been thinking a little bit about how to assign identifiers to new forms of scholarly communication- thinks like blogs, wikis, data sets, etc. Some background can be found here: http://tinyurl.com/m68jlq - Geoffrey Bilder

Yes, that's useful background. Do you have an update on the current situation? - Daniel Mietchen

hReview supports multiple rating dimensions if the site could use it - Mike Chelen

Google are about to roll out what they call "Rich snippets" for selected sites, harvesting info provided via hreview, RDFa or similar: http://bit.ly/3bGVlE & http://bit.ly/SxWES . Would be nice to see this coupled (for scientists logged in with their author ID) with ratings like at PLoS ONE. - Daniel Mietchen

Microformats for biological (and possibly even other, e.g. chemical) species ( http://microformats.org/wiki... ): "Imagine viewing a web page with a reference to a species - and being able to use an add-on to you browser to be taken directly to information about that species, on, say, Wikipedia, or Wikispecies, or Google Images, or another site, such as in an academic database, of your choosing." - Daniel Mietchen

@Daniel - check out NameLink from Enc of Life - exactly what you are talking about. http://labs.eol.org/... - Peter Binfield

Thanks, Peter: Yes, EoL was in the mind of the writer of this phrase (he also quotes Wilson), but NameTag and the species microformat are two different approaches to this goal, and only the latter bears some resemblance to article-level metrics. - Daniel Mietchen

A similar project: http://project.guiguan.net/ScoreFi... . - Daniel Mietchen

Interesting discussion going on at http://groups.google.com/group... . - Daniel Mietchen

Wonder what might be a good basis for building a similar service? - Mike Chelen