Jimmy Lin, Michael DiCuccio, Vahan Grigoryan, and W. John Wilbur.
Exploring the Effectiveness of Related Article Search in PubMed.
[Free full-text PDF here]
We describe two complementary studies that explore the effectiveness of related article search in PubMed. The first attempts to characterize the topological properties of document networks that are implicitly defined by this capability. The second focuses on analysis of PubMed query logs to gain an understanding of real user behavior. Combined evidence suggests that related article search is both a useful and often exploited feature in PubMed.
[Via the constantly-amazing Resource Shelf]
There’s been another big surge in online discussion of UpToDate recently.
Prediction: As time passes, UpToDate will have more and more well-designed, well-maintained, effectively-marketed, affordable competitors. The competition will bring subscription costs down to a more reasonable level.
To keep up with future commentary on UpToDate from the medical blogosphere, subscribe to this MedWorm feed.
When I was growing up in Rochester, Minnesota, we didn’t refer to the institution as “The Mayo Clinic.” It was just “The Clinic.” It would be an understatement to say that Mayo is an incredibly huge presence in my home town. When I asked one Rochester local if anything significant had changed in Rochester since my last visit, the local replied that Mayo had “completed its acquisition of downtown.”
After Melissa Rethlefsen escorted me to the venue for the presentation on blogs that I was invited to give for Mayo Libraries 2.0, she and I went out to lunch with MD/PhD student Colin Segovis. The Mayo School of Medicine accepts only 42 students per year, so students there tend to be particularly brilliant- and Colin is no exception. In addition to his medical studies, Colin is a great big geek (and I mean that in the best possible way). Melissa tolerated us both very well and actually looked amused as we degenerated into a sort of constant, geeky buzz.
Above: Mayo MD/PhD Student Colin Segovis and Mayo Librarian Melissa Rethlefsen
As Melissa took me on a tour of Mayo’s many excellent libraries (this took hours and we still didn’t get to a couple of them), I found myself constantly impressed by Mayo’s art and architecture (none of which I’d be able to appreciate without the years of education from my wife, a historian of art and design).
I didn’t take any photos of the Warhols.
Melissa’s (medical school) library is lovely. Note the dignified portrait that looks down on those studying below.
Above: Mayo School of Medicine Learning Resource Center
Mayo’s Cancer Education Library:
Plummer Hall Windows:
Click here for more (and better) photos of Mayo’s wonderful libraries
I regret that I failed to take a photo of the beautiful patient library where Susan Mayer works, but it sure was a pleasure to finally meet her. I wish our hospital could provide the kinds of materials and services her library provides to patients and families every day.
I also got to meet Mayo’s head reference librarian, Pat Erwin. Pat is clearly brilliant (and seriously respected by her colleagues) and awfully nice. We talked mostly about cats, a shared interest.
Debbie Fuehrer (who I have known since I was 14) runs a great patient library of mostly non-clinical materials including recreational reading in multiple languages, movies, and video games. I imagine it is a lot like running a very, very good (but very, very small) public library. It was loads of fun to see Debbie.
Melissa also took me to meet David Brown, head of systems for Mayo Libraries. I liked him immediately because of the huge collection of novels he kept in his office (which, he told me, is a tiny sub-section of his collection). Is it just me, or are all library geeks fans of fantasy or sci-fi (or both)? I really should’ve taken the opportunity to ask questions about what is involved in running Mayo Libraries’ systems, but we talked about things we like reading instead (which reminds me I need to get something to David Brown that I promised to send…)
Of course, the best part of my trip was finally getting to meet Melissa Rethlefsen in person after many months of emailing.
From the Microarray Lab, Department of Psychiatry / Molecular and Behavioral Neuroscience Institute at the University of Michigan comes PubViz, a powerfull, flexible prototype interface (built in Flash!) for MEDLINE information retrieval.
In short, PubViz is developed to provide the capability of utilizing external knowledge as well as interactive visual query functions for more efficient exploration of the Medline database. The current version has the ability to utilize protein-protein interaction data during Medline search and enable researchers to identify functionally related Medline records not retrievable in existing search engines. It can also utilize the structure relationship of different type of genetic markers including cytobands, microsatellite/STS markers, SNPs and genes derived from human genome assembly and HapMap data for deep search of genetically related Medline records. We include many visualization functions in PubViz, such as interactive PMID, MeSH, Gene views, the transition between different views, selection of node description display on network graph, as well as details of abstract and sorting/filtering functions. The combination of these novel capabilities will make PubViz a powerful tool for Medline exploration.
The site notes that PubViz currently searches only the sample corpus, but that they “plan to release full Medline search engine on 8/1/2007.”
I’m excited about the work because it is an enjoyable challenge and a great motivator to routinely write new, original material.
My first post was on 7/9/07 and it is my goal to write a post every week to help de-mystify technology topics and terms for libraryfolk of all sorts in Central New York. The CLRC blog content, therefore, is more general and more basic than the stuff covered at davidrothman.net, but if the idea of this appeals to you, you might wish to subscribe to the CLRC blog feed.
If you work in a medical library and haven’t yet subscribed to Mark Rabnett’s Shelved in the W’s, you probably will after reading this post.
Still catching up and discovering neat stuff in my aggregator from my time off last week, including this interview with T. Scott Plutchak at DegreeTutor.com.
Do you see academic medical libraries being a vanguard to other libraries and their adoptions of technology?
Yes and no. Yes in that I think that there are a lot of very smart, tech-savvy people in medical libraries; some of the angst that I see reflected on library blogs, from libraries in other sectors, reflects frustration from what they see as a slowness in adopting technology. I just don’t see that in academic medical libraries…
Go read the whole thing. While you’re at it, consider subscribing to Scott’s blog. If you’re like me, you’ll not only be impressed with Scott’s knowledge and insight- you’ll also envy the elegance of his writing.
Naturally, it is in the week I take a vacation from blogging that two really interesting new MedLib blogs come into existence. I subscribed to each as soon as it came to my attention.
Mark Funk (bio) has started the first-ever blog by an MLA president: Only Connect!. I’ve previously mentioned Mark’s enthusiasm for new technologies and I’m thrilled he’s blogging. No doubt you’ll want to subscribe to the feed.
Mark Rabnett (previously mentioned here and here) has started Shelved in the W’s: Working notes of a hospital librarian. Shelved in the W’s is off to an outstanding start and has quickly become one of my favorites.
One question for both Marks:
Many thanks to Mayo Clinic Libraries for having me- I had a great time!
(I’ll share more about my visit to Mayo in a future post.)
Just got back to NY and am starting the long process of catching up on email and reading. If you’ve written in the last week, please forgive the delay and know that I’ll get back to you in the next few days. There will be very light posting this week while I catch up both at work and at home.
Thank you to the wonderful Kaura Gale for keeping the lights on at davidrothman.net while I was away! I *will* need to take some time off the blog in coming months and would welcome email from anyone who would like to take a shot at writing a guest post (or a whole bunch of them).
Posts to expect later this week after I’m caught up: Notes on my visit to Mayo Libraries (short preview: Wow!), slides from my presentation for Mayo Libraries 2.0, multiple clumsy superlatives and metaphors which will attempt to describe how awesome Mayo’s Melissa Rethlefsen is (short preview: She’s crazy awesome), some articles of potential interest and a handful of resources online that have recently come to my attention.
Yesterday, commenter Iskandar informed me that in Italy, they say “Mamma mia” when suffering from information overload.
Mamma mia! Mamma mia! Mamma mia!
I promised I would discuss PubWindows today, so I’ll try. Honestly, I still haven’t completely figured it all out. Partly because of time, and partly because there is soo much to it. If anyone knows/understands more about it, I invite you to leave a comment; I think others will greatly benefit from your knowledge.
So, PubWindows appears to be the brainchild of the director of Neurotransmitter.net, Shawn Thomas, who has a MLIS. Cool! From the site: “Neurotransmitter.net exists to improve online access to biomedical information in fields such as neuroscience, pharmacology, and psychology… [PubWindows] allows you to browse PubMed citations with a modified version of the National Library of Medicine’s MeSH Browser. PubWindows offers search links to the On-line Medical Dictionary and PubMed text mining tools such as Chilibot and XplorMed . In addition, links to RSS feeds and various export formats are available.”
You can select a MeSH term from PubWindows and then use it to search XplorMed, which shows the association between terms in Medline abstracts, and you can also search SwissProt, SpTrembl, OMIM and SMART. I think. Mamma mia!
Or, you can use your term to search chilibot, which allows users to “[search] PubMed literature database (abstracts) about specific relationships between proteins, genes, or keywords. The results are returned as a graph”. The example graphs provided were really very visually easy to read and I could see how they’d be quite useful.
In PubWindows itself you can view the abstracts attached to your MeSH term and create a RSS feed for new ones.
I think PubWindows is really great idea, but I found it very difficult to use. I don’t know about PubWindows users, but I usually need to search MeSH before I’m able to select the appropriate term. Also, I wasn’t able to successfully search with subheadings, which I think is a huge drawback in terms of precision.
And as Iskandar notes, you can’t navigate back to the main PubWindow page once you’ve gone into one of the other databases.
I’ve been trying to come up with something witty about this being a neuroscience engine and using MeSH not needing to be brain surgery, but its late and I’m tired.
I wanted to write about PubWindows today, but I simply haven’t had enough time to evaluate it properly. I’ll post about that tomorrow.
In the meantime, I want to share with you a found haiku from the ‘introduction’ section in XplorMed.
cases in which you do not
know exactly what
you are expecting to find.
Your interests may change
by the results obtained,
you may want to ask
new questions as analysis
develops. Also, the
results may suggest you use
that should be used to expand
your MEDLINE query
(such as, an unexpected
of a protein name, or the
synonyms of a
disease or other problem).
slightly modified to fit form; form modified to fit length!
You really should click over and check it out. The design is deceptively simple, lovely and easy to use.
There’s not enough room for everything, so here are the highlights:
Hot Topics option:
Displays visual representations of how many articles with your concept have been published, top authors and journals, and provides a graphic of countries in which the research is being done. Notably, a disclaimer is posted that states quantity is not the same as quality.
Lists citations clearly, beautifully. Your search term (and impressively, the main topics of the article) are unobtrusively but obviously highlighted. The icons are clear and easy to use, including an option that lets you toggle between citation and abstract view for each listing. Also gives one click options for related genes, which is great, but I’m less (read: Not) pleased with the one click option for wikipedia entries.
What I like best is that on the left hand side, both GO and MeSH relevant terms are listed first in top popularity, then in hierarchy of content, and you can click on a term to refine your search. PubMed, take note!
Gives the ability to search in full text, anywhere, abstract or title, title, author or affiliation (and a few others), allowing for a lot of flexibility and focus. The simplicity of this page puts Google’s Scholar advanced search to shame.
Depending on the researcher and their topic, this is one of the few free third party tools I’d recommend.
I was going to review Twease today, but I realized that David already did a great job with it, and that one of its creators, Fabien Campagne, left additional notes in the comments section that make for a complete review of what it does and how it works.
What neither of them touched on, though, was how beautifully designed it is. Go take a look. Meant as a compliment, I’d call it the “targetization” of Medline!
Also, I think it has one of the most straight forward, easy to use and understand tutorials I’ve seen.
And while I’m on tutorials I’d recommend, check out this Ovid Medline one from Duke.
Earlier I was thinking about which third-party PubMed tool to review, when I noticed that something didn’t seem right. Then it hit me: I’m not being specific. I’ve been using the terms PubMed and Medline interchangeably, when that is incorrect.
MEDLINE is the largest component of PubMed…snip…In addition to MEDLINE citations, PubMed also contains:
In-process citations which provide a record for an article before it is indexed with MeSH and added to MEDLINE or converted to out-of-scope status.
Citations that precede the date that a journal was selected for MEDLINE indexing (when supplied electronically by the publisher).
Some OLDMEDLINE citations that have not yet been updated with current vocabulary and converted to MEDLINE status.
Citations to articles that are out-of-scope (e.g., covering plate tectonics or astrophysics) from certain MEDLINE journals, primarily general science and general chemistry journals, for which the life sciences articles are indexed with MeSH for MEDLINE.
Some life science journals that submit full text to PubMedCentral® and may not yet have been recommended for inclusion in MEDLINE although they have undergone a review by NLM, and some physics journals that were part of a prototype PubMed in the early to mid-1990’s.
I’ve been saying PubMed third party tools, when what I *so far* have meant is third party tools for Medline.
So, today I tried out Medie. I’d said I’d write about it as a third party Medline tool. I can’t. At least not as a hospital librarian. I was going to try to give it a pat on its back and insincerely flash it half a smile so it wouldn’t feel too badly about itself.
Instead, I’ll tell you what I really think of it. The first problem I noticed was that it doesn’t use MeSH (and therefore you can’t focus or use subheadings), so you’re losing massive precision right away. The other biggie I couldn’t get past was that it lists by PMID number, not title and author.
Sounds like I hate it, right? Not at all. Once I learned more about it and understood what it does, I was blown away.
This has changed my mind completely.
Let me take a step back for a moment. One of the first things I mention when I teach Medline is that you are using a computer. You can’t talk to it like you are asking me a question. You need to represent your concepts with words and terms and phrases and connect them appropriately. Then I launch into Boolean logic and MeSH.
Medie is a project of Tsujii Laboratory at the University of Tokyo that works on Natural Language Processing and Computational Linguistics. Basically, this means that it can root your word and process algorithms so that you search the database with natural language. If it doesn’t already, it will be “thinking” that when you type heart you may also like results with cardiac. That is too basic of an example, but hopefully you get the idea.
So, when you search Medie you enter search terms into basic parts of speech (subject, verb, object) and out pop your results, with the line or two of text it picked up color-coded by part of speech. It does have some additional search options that I can see as being very useful and helpful. But as I can’t recommend it for searching now, I’ll let you discover those on your own. I should note that I didn’t find any information on what text is searched or omitted, or how results are ranked.
But in the future I can see Hal saying: “I’m sorry, Dave I’m afraid I can do that.”