CiteULike

British Medical Journal (BMJ)

Wed, 29 Jun 2005 13:26:37 GMT

You can now post articles directly from the British Medical Journal to CiteULike. Many thanks to Will Wade for writing this plugin using the developer's kit. I wonder if whoever eventually works out a cure for cancer will be a CiteULike user. That would be - as my American friends say - "neat".

Anthrosource

Mon, 27 Jun 2005 21:58:54 GMT

You can now post anthropology articles from anthrosource to CiteULike using the bookmarklet. This is another user submitted plugin using the plugin development kit. It turned out that two people were independently working on an anthrosource plugin, but Christopher Kelty got there first. I never knew there were so many anthropologists in the world...

Update: It seems there are, indeed, plenty of anthropologists in the world, and they've written an exceptionally comprehensive introduction to using CiteULike with Anthrosource. It probably works just as well if you want to use CiteULike with any other source of articles (particle physics, for example), and I'm actually embarrassed that they've written a better tutorial than I have on the site itself...

Smithsonian/NASA ADS Astronomy/Planetary Abstract Service

Sat, 25 Jun 2005 01:07:27 GMT

The CiteULike bookmarklet now works with the Smithsonian/NASA ADS Astronomy/Planetary Abstract Service.

API for developing plugins released

Fri, 24 Jun 2005 19:02:52 GMT

I've released an open source version of CiteULike's plugin interface (the code which scrapes citation details from external sites) under the BSD license. It's available to browse here:

http://svn.citeulike.org/svn/plugins/

and the HOWTO which provides the documentation as it stands at the moment is here:

http://svn.citeulike.org/svn/plugins/HOWTO.txt

If you want to pull the source code down and start hacking it, you can get it using subversion (http://subversion.tigris.org/) by typing:

svn co http://svn.citeulike.org/svn/ citeulike

The relevant features are:

Language neutral. You can write plugins in whatever programming language you like (assuming it can run on my server).
Sample code. I've tidied up a few of the existing plugins and released them as part of the new system. Ultimately I'll convert them all, but it's now at a state where it makes sense to release it and get everyone hacking on it.
Proper documentation. It walks you through all the steps required to produce a new plugin.
Test suite. You can (and should) write test cases for your scraper, and we'll know when the site changes its format and breaks the scraper (such is the nature of writing these things).
Test harness. You can actually run the tests against your code without having to guess whether they'd work or not (which was the case up until now if you wanted to write a plugin).
Common utility functionality (Author names, RIS, and BibTeX parsing) build into the "driver" part of the code, so you don't need to re-invent the wheel.

It's live on the server now, as is the first user submitted plugin (from Diwaker Gupta) to scrape the proceedings from the computer science journals on the USENIX site.

Edit DOIs for your articles

Mon, 20 Jun 2005 18:13:12 GMT

Another new features lets you add a DOI (Document Object Identifier) to your article on CiteULike if it doesn't have one and you know what it should be. Just click down to the view article page, and hit the "Edit links and DOIs..." button. You'll obviously need to be logged in, and you'll need to have the article in your collection.

Apparently some journal styles actually require you to produce bibliographies with DOIs, so this should help if you publish in one of those.

MathSciNet Parser

Sun, 19 Jun 2005 16:45:59 GMT

I've added experimental support for MathSciNet. You should be able to use the "Post to CiteULike" bookmarklet to automatically add articles from MathSciNet to your library without all that tedious mucking about copying and pasting details.

By experimental, I mean that it seems to work, but I'd be interested to hear from you at bugs@citeulike.org if it's not behaving. To be trendy, I should probably call it Beta or something like that. Google seem to have recently redefined that word to mean just "new feature".

Find your Personal PDFs more easily

Sun, 19 Jun 2005 10:19:30 GMT

I'm now back from holiday. After a grand total of sixty seven hours sitting in railway carriages in Western Russia, I'm now surprisingly keen to get back to work.

The first minor update follows on from the Personal PDF announcement. When you upload your own copy of the PDF file, you'll be able to see a small icon in the list view to let you tell at a glance which files you've already uploaded and which ones you haven't. Previously, you had to go down to the "view article" page, which was a real pain.

Update: The same now applies to articles with notes. If an article has public notes attached to it (or it's your own article with private notes and you're logged in) then you'll get a handy icon in the list view too. It looks ugly as anything, but it's quite useful.

Japanese CiteULike

Fri, 27 May 2005 17:01:23 GMT

My thanks to Marica Odagaki for producing a Japanese translation of CiteULike. I'm always amazed by how much effort goes into translating an entire web site and how people like Marcia kindly volunteer to help out. Hopefully this will help with all the Japanese users currently use Google Translate to access CiteULike. I've no idea how well it translates into Japanese, but it has certainly produced some fairly interesting Japanese->English ones when I've used it.

Personal PDFs

Mon, 25 Apr 2005 07:59:24 GMT

Do you keep copies of the PDFs of some of your articles on your hard drive? Can you never find the right one when you want to? Or, do you keep these files on a machine at work and sometimes want access to them at home?

Now you can keep a personal copy of the PDF on the CiteULike server. When you're logged in, just navigate down to any article with your web browser and you'll get the option to upload the PDF from your hard drive. After that, you'll have access to the content wherever you can log in to CiteULike.

There are obviously some restrictions to comply with copyright law. In order to prove that you have access rights to the content, you must be able to upload the PDF file directly into your account. That's to say, you must prove that you had a copy of the article in the first place - CiteULike will not be able to get hold if it in any other way.

Secondly, rather obviously, you must be logged in to your CiteULike account to download the article again when you, say, want to read it at home. That's the "personal" aspect. Think of this service as (a more organised and convenient) extension of your local hard drive. You can keep your content on it, but you can't use it to share that content with anyone else.

All data is backed up to a remote site and, in the initial testing phase, there are no usage quotas. I'll clearly need to impose some sort of limit at some point, but I've got plenty of disk space and I want to get an idea of popularity so I can work out where I should peg the limit.

Coming next (once I'm happy I've set the usage limits correctly) is the ability to do a full text search on all the PDFs in your collection. This will hopefully solve the "where did I read that?" problem - all you'll need to do is ask CiteULike to search through all the papers you've ever read and tell you exactly where you read it.

Optimisations

Mon, 21 Mar 2005 13:27:08 GMT

I've rewritten a substantial chunk of what goes on behind the scenes on the server. You shouldn't notice any difference except from:

The "delete article" button has moved from the list page to the article page itself.

If two people post the same article in rapid succession, you only see the latter as opposed to filling the page up with multiple instances of it.

and..

page generation is 10 times faster than before (for some of the more commonly viewed pages).

The chances of this all working entirely smoothly are pretty slim, so please let me know if you notice anything untoward happening on the site.

The geeky explanation of what I've done is as follows:

I replaced a simple shared-memory hashtable with memcached together with some modifications to handle dependencies in data as well as a custom Tcl client API.

The advantage of this is that I can pre-cache the HTML rendering of each article in the list (hence I moved the "delete" button so the HTML doesn't depend on whether you're logged in or not), and use memcached's method for fetching multiple objects in one request to simply suck down all the HTML fragments and then simply play them out to the webserver. Not only is this approximately 10 times faster than what I had before (a typical page with all the articles in the cache renders in 12ms), but most of the remaining time is taken up just waiting for the data to appear over the network - and this consumes next to no CPU at all. That means I've managed to get at least an extra factor of ten capacity out of the existing hardware (and there are probably still some more optimisations I can do if I need to), and I can now use multiple instances of memcached to effectively scale up capacity by adding more hardware when required.

So, I'm now in a pretty good position to cope with growth of number of users (which is exponential), and I can now get back to spending more time adding features (like finally writing that API) than worrying about this stuff. The only thing I still need to do, of course, is fix all the bugs which I've just introduced - please do let me know if you notice anything broken.

Import your existing BibTeX file into CiteULike

Tue, 01 Mar 2005 00:21:28 GMT

Do you want to start using CiteULike but you've already spent ages making a BibTeX file packed with every article you've ever read? Fear not, you can now import it into CiteULike.

Amazon books

Sun, 27 Feb 2005 00:21:21 GMT

I've added an Amazon plugin so you should be able to post books from their web site using the normal bookmarklet. You can't post any of the other things they sell (like microwave ovens, for example) as that would be extremely confusing.

Roundup of two minor new features...

Tue, 15 Feb 2005 10:00:56 GMT

For the librarians: Install one of Dan Chudnov's magic dynamic appropriate resolver bookmarklets and click it when you're looking at an article on CiteULike. Dan's trying to push the ideas behind this as a standard for embedding this sort of information in web pages, and I'm all in favour of that.

For the fastidious: You've now got much more control over how you can edit the detail of your articles once they're in your library. Clicking the "Edit details..." button on the article page should let you do things that you previously couldn't do (like changing the article type, and adding abstracts), while clicking the "Edit links..." button will let you associate another URL with the article. This is handy if you eventually find a PDF of it online.

New language - Italian

Sat, 12 Feb 2005 15:00:07 GMT

... or, at least it's new to CiteULike, even if the language itself has been about for a while (something to do with the Romans, I believe). Many thanks to Paolo Massa for his hard work translating the site into Italian. There's barely room for the array of flags in the top right-hand corner of the page any more - I'm amazed at the sheer number we've got now. Thanks again to everyone who's contributed to all these translations.

citeulike-discuss mailing list

Fri, 04 Feb 2005 13:41:26 GMT

It's high time we had an email discussion list for CiteULike. I've set one up, and you can subscribe here:

http://www.citeulike.org/mailman/listinfo/citeulike-discuss

Feel free to use it for general discussion about the service, ideas for what you'd like CiteULike to do, and any questions or concerns you might have. If you've got a question about your account, or a specific bug report, that should go directly to me (Richard Cameron - camster@citeulike.org) and not to the list.

Extremely long lists of tags - filter them!

Wed, 02 Feb 2005 14:20:17 GMT

I've added a little filter box to the list of tags on the right hand side of the pages. I find this quite useful to find stuff now my list of tags has grown to a horrendous length.

It's a bit of a javascript hack, and it's possible that it might not behave itself on every browser, so I'd be interested to hear if it doesn't work out for you (unless, of course, you're using a text browser like Lynx!).

New plugin: Science

Mon, 31 Jan 2005 17:44:21 GMT

Science is now fully supported in CiteULike, which means you can add articles from it without having to type in the details.

IEEE Xplore

Fri, 28 Jan 2005 18:13:30 GMT

... and you can also post articles from IEEE Xplore and have the details automatically appear for you.

Nature

Fri, 28 Jan 2005 16:43:58 GMT

Quick update: You can now post articles directly from the nature.com website without having to type the details in manually.

Post anything - not just from supported sites

Tue, 18 Jan 2005 00:27:35 GMT

I've obviously been spending too much time working on CiteULike and not enough time using it. It's only in the last few days I've discovered just how frustrating it is to get an error message when CiteULike won't let me post a particular article because it's not from a supported site.

So, I've removed this restriction. You can now post any URL you like. The only downside is that if it's not from a site that CiteULike knows about, you'll have to type in the citation details yourself.

There's a bit of history to why I decided to impose the restriction in the first place. When I wrote the site I didn't want it to become yet another social bookmarking site. I wanted it to fill a specific niche - academic articles. For it to be a useful academic resource, I reasoned, I had to make sure that users could browse about and get academic articles and not, say, details of chalets in France posted by a biochemist trying to organize his skiing holiday. That sort of thing would seriously devalue the site.

Secondly, I wanted to make sure the citation details were reasonably accurate (and spam free), so I initially decided to only take this information from serious online databases (like PubMed and Sciencedirect).

However, it's just far too annoying to be limited to a finite set of providers, so I had to make the change. There's still the issue of not polluting the quality of articles with arbitrary web pages though. To solve this, if you post a web page (or a PDF file, or whatever) from an unrecognized source, it will appear in you library, but not on the front page. Of course if you post the same article as others you'll still be able to see the "posted by x others link" to find out who. This should hopefully be a reasonable compromise which will limit the damage that spammers can do to the site while allowing users the flexibility to actually store their entire bibliography online, and not just part of it.

Hope this helps.

Two small things: edit details and private notes

Mon, 17 Jan 2005 21:21:15 GMT

Two quick updates:

When you post a note on an article you've now got the option of making it private. Only you will be able to read it, and you'll only be able to read it when you're logged in. This means you can criticize your boss's work to your heart's content without him seeing it.

You can edit the metadata (author, title, journal name, page, etc) of articles in your collection. Sometimes (especially with Citeseer) the data which gets imported is just wrong, and you need to manually fix it up. This is really handy if you then want to export your library to BibTeX or EndNote. Your changes only apply to your copy of the article, so they won't appear on the front page of CiteULike or in anyone else's library.

Watchlists, now without that really annoying "feature"

Fri, 14 Jan 2005 14:16:27 GMT

I've just made a small change to the way watchlists work. It's something which has been annoying me for ages, but I've only just got round to fixing it. Previously, if someone posted an article which matched your watchlist criteria, then it would appear regardless of whether you already had it in your library or not. This seemed to happen quite often, and it only served to annoy me.

So, from now on, you'll only see articles in your watchlist which you haven't already posted into your collection.

CiteULike in Norwegian

Tue, 11 Jan 2005 12:41:58 GMT

There's now a Norwegian version of CiteULike. We seem to have a disproportionate number of Norwegian users on the site, so hopefully this will be useful for them.

My usual plug for anyone else interested in translating CiteULike (or even part of it) into their native language still stands! I'd be delighted to hear from you at camster@citeulike.org.

How much do you want to read this article?

Sun, 09 Jan 2005 00:36:22 GMT

My paper-based filing system (which has now been almost completely replaced by CiteULike) was an enormous pile of papers, all wrapped up in a folder with the words "to read" written on it in big letters. I also had some general procedure whereby the articles that I really wanted to read got inserted somewhere near the top, and the ones I wasn't really in a hurry to get through went somewhere towards the bottom.

I've added this functionality into CiteULike, so when you post a paper you'll get asked how urgently you want to read it. Your options are:

Top priority!
I really want to read it
I will read it
I might read it
I don't really want to read it
I've already read it

Afterwards, you'll be able to order your library in the way my paper file used to work - things which you ought to read first at the top of the list. Obviously, you can change the priority via the "view article" page when you're logged in too.

I decided against letting users rate papers in terms of quality, and just went with recording how high up the user's personal "to do" list the article is. There were a few reasons for this:

In my paper-based world, when I finished reading an article, took it out of the "to read" folder and put it into its permanent home in the filing cabinet, I'd never in a million years think of scribbling a mark out of ten on it. Firstly it would sometimes be incredibly presumptuous to do so, and secondly it doesn't really achieve anything. If I realize that the article is of no interest to me and that I'm never going to cite it or refer to it again, it just goes straight in the bin; otherwise I keep it. It's really just a simple yes/no choice for me.

However I did find myself instinctively prioritizing all the stuff in the "to read" folder, and so I think that's the obvious ranking to use on CiteULike. It also avoids any potentially embarrassing situations where the author of an article (especially if he happens to be my boss) gets offended that I've given his article a bad mark. I haven't. It's just that it's not strictly relevant to what I'm reading at the moment.

CiteULike is now searchable...

Fri, 31 Dec 2004 21:24:08 GMT

I've added a search bar to the top of every page. Now that we've got something like 70,000 articles online, it seemed sensible to actually be able search the whole collection.

I use a fairly standard approach at the moment: stemming, stop words, and ranking search results by: a) how close together the search terms appear in the record, and b) the idea that matching an author is better than matching a title, which is in turn better than matching in an abstract, which is better than matching anything else (like a journal name or a page number).

There's a lot of room for improvement here. The grand plan is to try to make the search results more relevant to you. For example if you're an astrophysicist (and have posted lots of articles on astrophysics to your library) then it's probably a safe bet that you're more interested in results from astrophysics journals than you are in sociology journals. This is very much the grand plan, but I'm sure we'll get there eventually.

Happy new year!