The Hanover Collaboration

“Search don’t sort.” Say that three times. This is the new world. Managing large amounts of data by hand is a fool’s errand.

I look at a lot of papers in a day. I mostly don’t read them as much as take note of them. Then I want to file the thing away along with a few comments.

In the old days, your workflow perhaps worked as follows. You would download the thing, file it in a directory tree, and put your notes in some huge text file. When you go to write the paper, you start from the notes, dig out the file and then make a citation key for it. This is a fair bit of work. Worse still, only the comments are searchable.

Smarter people started using Google Desktop or spotlight to make the documents searchable. Those of us who know that Adobe isn’t the only PDF program, probably figured out how to attach comments directly to the files with a tool like Skim. This amounted to a decently workable solution. Two problems though: 1) spotlight and google desktop are not specific enough of tools. (e.g. you might want to search the abstracts only, or filter for year) 2) you still have to manage the files and citations yourself.

There are a number of tools called “citation managers” that do one or both of these things. Up until fairly recently, all the good ones cost money. Nowadays, Bibdesk is considered quite adequate by most… Most Mac users that is, it doesn’t run on Windows. It will organize the files, index their contents and metadata and make everything nice and searchable. It also indexes Skim comments as they are added to the documents. It is kind of a pain getting the references put in in the first place though. I found myself writing an unacceptable amount of bibtex by hand.

There is also a Firefox extension called Zotero that will manage citations directly from your browser. You can automatically add search results from Google Scholar, Citeseer and a number of library website formats to a bibtex file, complete with a link and optionally, a full abstract. This takes one or two clicks.

It turns out that you can actually use both. You can use Zotero as an input to Bibdesk. This is just short of perfect. You still need to actually download the paper (i.e. right click on it and save it somewhere) if you want to get the full text indexed by Bibdesk. Anyway, next time you are tempted to “clean up” your references, you should consider using some tools like these.