Connontea
Connotea.org is a collaborative tagging system created by Nature, inc., mostly for bookmarking academic articles. It was mentioned in a recent paper about collaborative tagging systems, along with deli.cio.us and CiteUlike. The similarities and differences among three systems are quite interesting.
Connotea and citeULike are both specifically designed for annotating academic papers, and deli.cio.us is a general purpose social software for web-bookmarking. However, Connotea is more similar to deli.cio.us than to Citeulike.
- citeulike is a standalone bibilographic system, with the ability to (a) extra bibilo info from various publishers’ webpages and (b) allow collaborative tagging. The Citeulike site stores the bibilographic info, including the abstract, which I am not sure if it is copyrighted. Connotea, on the other hand, is literally a bookmarking system. The DOI is the primary identification info.
- Related, you can manually post detailed biblio info to citeulike (but curiously missing are DOI and ISSN fields). In connotea, you can only post an URL, and a title. There is no way to add the full reference, as far as I can see (one may develop a plug-in for that purpose, but I am not sure the dababase has fields for authors, journal, etc.
- CiteULike is very smart in taking advantage of the OpenURL, or Z3988 links. It constructs the Z3988 string from the biblio info in the database, if it’s not directly available at the time of posting (Is this right? Got to double checking!!). Interestingly, it embeds Z3988 as a <SPAN> element in the HTML, it relies on DOI and the publisher URI for resoving the paper (solutions based on GreaseMonkey are available). Connotea creates an OpenURL resolver URL for you if you give it the SFX resolver URL from your institution. But you can’t do that if the entry is manually posted. There is simply no such info in the database.
- Every entry in Citeulike is public — which makes sense because these are supposed to be "publications". The only thing that can be private is your comments. In contrast, Connotea allows you to post an entry as public, private to a group, or private to yourself. Again, this reveals its origin in webpage bookmarking.
- Both systems have "groups," although both get it wrong. In the current systems, anything you post will show up in all the groups you "belong" to (Connotea allows you to make an entry private, which means nobody else can see it). That’s not how I would use a group function. The core of a group is a shared interest on a topic area, but a use may have many interests. It makes no sense to share papers on philosophy with an interest group in phonology just because I happen to be interested in both. Connotea’s approach is problematic because I should still be able to share my philosophy paper with my philosophy friends, and shouldn’t be forced to make it private just because I don’t want to share it with my linguistic friends.
I’ve said it elsewhere: The only sensible way to organize a group in the academic setting is to allow users to specify whether they want to share — in the sense of recommend — this paper with a particular group. This can be done with a combination of default settings and manual correction, or a tagging-based method.
- To me, the biggest attraction of Connotea is that it’s GNU. CiteULike is free but not open, and its future is unknown.
The better system would be a combination of the strength of both citeulike and connotea:
- open source
- full biblio database <– copyright?
- separation of the biblio database functions and collaborative functions
- biblio db: full API support, including posting, searching, editing, and duplication/error checking (self-maintaince)
- collaborative tagging and commenting: wiki-like pages for each entry.
- web-interface: AJAX-based.