Sunday, May 11, 2008

Uses for RSS

I am still amazed by how few people in academia use RSS. Seriously, people, get with the program. The other day I was sitting in a meeting and the PI asks, "Hey, what's (one of those labs we always watch for new papers) latest paper?" I pop up Google Reader, hit my PubMed feed for that search and start rambling off the last 5 papers. So, though I mentioned it before, here is an updated primer on RSS use in research.

First, though, why use RSS?
- You never have to navigate confusing journal sites, deal with their crappy search engines.
- You are instantly updated on all info in the field without having to do anything beyond the initial quick setup.
- All data is portable. Don't like Google Reader? Switch to another RSS reader and bring your feeds with you.
- You can build in your own way of prioritizing topics. Feeds take up no room, they don't care if you look at them every day, or once per year.
- Sharing info is remarkably easy. Sharing queries allows you to keep the whole lab on the same page. Feeds can be shared, posted on web pages, given to new lab members, created on the fly, and many other aspects that make collaborating easier.
- Archiving easily. You can mark the stories you like, and then create a feed out of those. You want the lab to read a specific paper for lab meeting or journal club each week? Just mark it. For example, I don't have everything set up perfect right now, but if you want a sneak peek at the stories I will discuss on the next Digital Trends Podcast, here's the feed.


PubMed
PubMed has the option to create custom feeds for any query. Here is how I have it set up, and highly recommend it. There are 3 phases to PubMed RSS optimization:

1) General Topic Tracking
PubMed's sloppy search engine can annoying. Say you want to keep up with motor cortex papers. Searching "motor cortex" returns anything with "motor" and "cortex", which includes molecular papers about protein expression, etc. That's okay. You could always learn a thing or two from an off-topic paper. AND remember that you are not obligated to read ANYTHING in the feed. You can tap the "Mark all as read" button and poof, they're gone.
I use several general searches that I go through last when checking my feeds. I think they're still valuable to have, since every once in a while they catch a gem and might stimulate the creatively looking at questions from multiple angles.

2) Specific Topic Tracking
Alright, the nitty gritty of PubMed. No, don't bother to learn all the syntax. Just hit the "Limits" tab, which lets you set up filters. Here is where you will spend most of your time. If you have played around with the general search, and have around 5-10 important, recent papers that you know were essential, start tweaking your search so that all of those papers show up in the results. Don't confine yourself to one search, either. You might have papers on, say, motor cortex and movement trajectory decoding, and you might have another set on LFP activity in any cortical area. Fine. You have two feeds you want to create. One "Motor Cortex Decoding" and one "Cortical LFP". Don't be afraid to make too many feeds, since you are looking to make this section as specific as possible, but also as accurate.
Yes, PubMed can suck if you don't know the details of how it searches, but you can take the time to do a very specific search with the correct wording because it will be the last time you ever use that search again.

3) Important Author/Lab Tracking
Of all the searches I do, this has been the most useful. You know the labs you want to watch. Note the most important folks (PIs, and any collaborators). Search for each set of specific authors (1 per lab), so something like "Donoghue JP[Auth] OR Hochberg LR[Auth]" (no quotes). Create a feed for each.


Individual journals
Almost all journals have the option to subscribe via RSS. You can keep up with the articles if you want, or just have the feed handy rather than having to navigate through the journal's site. "Hey, did you see that paper by Monosov in Neuron last month?" I just went from this browser window for this post to the answer in two clicks. THAT is powerful.


Individual sites
Like DNI, Digital Trends, TechRIVET, Ben's Bargains, CNet News, BoingBoing, and MetaFilter? Check them every day? Wouldn't it be easier if they all appeared in one place and you never had to hit up the site if there wasn't anything new? RSS readers will tell you what's new, let you categorize your feeds, flag stories with keywords, and make every aspect of your browsing easier.


Coordinating with others
In Google Reader at least, there is the option to share stories and attach notes about them. "Hey, check out the methods section. I think we could use that for our blahblah work." No coding up html, copying and pasting, emailing, etc.


Offline and online and via email
There are many RSS readers, both in standalone and online format. I'm not going to review them here, but I like Google Reader, since the Google Gears component allows you to take it offline as well. You can also have your feeds sent to your inbox using Feedburner or RSSFWD. Outlook 2007 has built in RSS support as well.


I went from 2-3 hours of active browsing per day for tech and research to around 10 minutes. I have returned to around 2 hours of browsing per day, but now I read around 1300 stories/day (by "read" I mean read the title, decide if I want to read the story, etc. I actually read about 20-25 full actual stories per day.) For example, I haven't hit my "General" folder yet today...
Anyhow, back to work...

No comments:

Post a Comment