What I want from a citation manager

  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  

Stockholm City Library. Photo (cc) Dick Rochester
Stockholm City Library. Photo (cc) Dick Rochester.

As a response to Alan Cann’s comments on Mendeley, I thought to put up general thoughts about what I want from a citation manager. I don’t have as clear an idea about what I want as Alan, so I’ll have to work it out as I go along.

Firstly and most importantly I want it to manage citations in documents as seamlessly as possible.

The main reason for using a citation manager for me is to wrangle the bibliographic details. These will go into documents or, occasionally, blog posts. Word 2008 for Mac has a bibliographic manager and it does the job, so any extra software has to offer something more, and also work well with Word. It’s a high bar, but not impossible because the bibliographic management in Word isn’t that great.

The surprise for me is that none of the online citation managers have an online word processor built into their sites. Given the number of online word processors and text editors I thought this was a marketing problem. Mark Grimshaw’s Wikindx is an excellent piece of software, but it’s a word processor written by someone whose primary job is something else. I thought professional programmers could solve this problem. I’ve been told they can’t – which makes Mark Grimshaw’s work even more impressive. The Zotero team has the idea almost right. You can do work in the browser, so a browser-based citation manager makes sense, but then so too would an integrated browser-based word processor. I’ve tried WizFolio with no success, but your experience may be different.

I’ve looked at Steve Nacin’s Simple Footnotes plug-in for WordPress. I think that could be modified with the use of custom taxonomies to work as the basis for an online citation manager. If that’s the case then another feature I need from a service is portability of data. Again, Microsoft Word is a good example. There’s .docx and .doc for saving, but there’s also .rtf and .txt. The online services do all have BibTeX export, but if you’ve tried using it you’ll see that BibTeX isn’t a simple standard. JSON makes sense. So too would RSS. Like Alan I like RSS everywhere. It might not be the powerful format, but it is simple which makes it useful.

I also want to see details about a paper quickly. This is somewhere where Mendeley currently falls. If I want to see the page for a paper there then from my dashboard at www.mendeley.com I:

  1. Click on Library
  2. Click on the collection a paper is in (don’t click on a paper’s title, that’ll give you the abstract, edit tags and edit notes)
  3. Click on the small text just beneath the collection title (don’t click on a paper’s title, that’ll give you the abstract, edit tags and edit notes)
  4. Then click on a paper’s title.

You’ll still get the abstract, but you’ll also get the related papers in the right column. Looking at the list above, the obvious extra step is having to click on a collection’s title to get to the collection page (step 3) after already selecting the collection (step 2). If I’m missing something, like a link I’ve overlooked, please let me know. I’ve also just realised I have no idea how to get to a paper’s web page from a shared collection. This is not a help if you’re interested in exploring the social features, and an online service without social features is going to be competing with desktop software purely on integration. Though I’ll also concede this might not be entirely fair. On CiteULike, I just type the tag I’m looking for into the search box. Doing the same in Mendeley does get me to search results where I can click to see paper details.

Contrary to Alan I think Mendeley’s use of a desktop companion might be a good idea. A lot of researchers don’t see social features as a boon, at least not immediately. PDF management is a help. Integration with Office is essential. The big competitors for all the online bibliographic sites aren’t each other. It’s EndNote. Thomson Reuters has recognised they are a threat and responded with a law suit against Zotero. Alan has said that EndNote is overpriced he may be right but a lot of that fee will go towards marketing you the next expensive upgrade. That’s a serious disadvantage for any competitors and having a desktop client is a way of making the change seem as if it’s like-for-like. It’s also reassuring. A web service might disappear overnight taking everything with it, while a desktop app sits securely on your hard drive. This means nothing if your hard-drive gets fried one morning but for some reason people don’t think so much about that. It might not be so useful to him, for good reasons, but if online managers are going to get traction then they’ll need to appeal to people who are reluctantly on Windows XP, because they were forced to upgrade from Windows 98.

Price is important and I agree with Alan. Compared to EndNote and RefWorks, £5 a month is reasonable. While I think that in my head, I paid around $25 a year for Flickr and Picnik. That’s seems to be my level. On the other hand I’m not sure I really need the extra features £5 a month buys on Mendeley. That leaves me at free, which I can’t complain about.

The biggest complaint I have about all the online services is that there are very few people on them in my field. I think a major reason Alan prefers CiteULike is that that’s where his network is. I’ve not yet got a useful network effect on any system. For all the various technical improvements that could be made, the winners will be the ones that work out the social problem of getting participation. That’s a marketing problem.

So how do you get people to sign up without it feeling that they’re aggressively being sold something or coerced?


  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  •  
  • Hi Alun!

    I can definitely see where your confusion comes in looking for the paper’s page in the research catalog from your library. I don’t think it’s intuitively obvious for people who are used to catalog interfaces (i.e. most researchers) to not go the abstract page for the item when you click on the title. There’s a division between the user’s individual private library and the collective public library that isn’t made really clear in how the web UI is designed. We could possibly address that by adding a link that points to the canonical document in the research catalog to every entry.

    I’d like to hear more about the features you would value and what they’d be worth to you. Integration with Google Docs has been discussed, but the API still has some issues and we haven’t been able to put the time into this that it would require. Maybe someone in the community could take a stab at this using the Mendeley and Google APIs?

  • Egon, the workaround there is to make a public collection and call it “My Publications” (or “Publications of Egon Willighagen”), then put all your publications in it.

    FYI – We are reworking how these features work for the web release on Wednesday.

  • I think you’ve hit on my key problem, Mr. Gunn. I’m not always clear whether I’m in a public or private space on Mendeley. I think the preference to keep data private is the best way to organise work, even though personally I’d be ok with public as default. I don’t think that would be a popular choice with most people. Some clear signing between public and private pages could help.

    I think integration with Google Docs would be the biggest attraction for me and I’d happily subscribe for that. I’d guess the next request after that would be for more citation styles.

    Another feature I’ve realised is missing is look-up by ISBN. Not a problem for the sciences, but it would be a time-saver for the Humanities where monographs matter. LibraryThing has this, but I don’t know how difficult it is to add.

    I’m also wondering if corporate accounts would be a good idea. I’ll have to think more about what I would want from an AoB point of view that I don’t need as a researcher.

  • WizCite is WizFolio’s citation tool that will work with online word processor as well as desktop word processor. This citation tool is not only compatible across platform but is also format agnostic. WizCite will work in most web based word processor such as Google Docs, Google Site, WordPress and Zoho Writer. However after working for months on the citation engine and just as we were ready to release in June, Google released their completely rebuilt Google Docs rendering WizCite incompatible. However you still have the option of using the old version of Google Docs in which WizCite will work. We are contacting Google Docs to make the necessary changes.

    Because WizCite is truly cross platform and format agnostic, it allows true collaborative citing across disparate word processors. For example you can cite in Google Docs (version prior to June 2010) and then export it as a Word Document file. Your collaborator can continue to cite the document in MS Word. Your collaborator can then send the updated document to another collaborator who will be able to open the document in OpenOffice on a Mac and continue citing the document without missing a beat. All this can be done without the need to send along a “library” file.

  • Thanks, Alun. We are actually doing a pilot of corporate accounts with some US-based universities now, so I’ll have someone get in touch with you about that. You’re not the first person to mention the Google Docs or the ISBN features, but we’d want to make sure we work with them so changes to their API don’t break our integration.