Frequently Asked Questions
Note: To preserve sanity, the news pieces are called "stories" rather than "articles", which I use only to refer to
scientific papers.
How do you identify the correct article for a given news story?
The software looks at the text of the story to identify certain features:
- Journal names
- Personal names (potentially those of authors)
- Institutions (that might be those of the authors)
- Dates (to get an idea of when the paper might have been published).
It uses these features to get a set of papers that might be relevant. This set is then ranked by similarity between
the abstract and the story.
Where do the papers come from?
They come from MEDLINE, an index of
peer-reviewed biological and medical articles published by the National Library of
Medicine in the US. It is available for anyone to search through PubMed.
Why is no article found for this story?
Several reasons are possible:
- It's a very recent article and hasn't made it into MEDLINE yet. It may be available on the publisher's
website; this is probably the best place to look if the journal is named in the story.
-
It's a story based on proceedings of a conference, a press release, a government statement, or some other source
that's not a journal article.
- It's based on an article in a journal that isn't indexed by MEDLINE.
- The story doesn't give us sufficient
information to identify the article - or the software here has missed out on
identifying the relevant bits of information.
Why can't I see the full version of an article? What do I do now?
Not all publishers allow people to view their content (especially recent content) for free. Some allow you to
purchase an electronic copy of a single article, usually for a few tens of dollars. Otherwise, your best courses of
action are probably:
- Search the web for the title of the paper. Sometimes the authors put up a copy of the material on their own
websites.
- Most universities and some public libraries will have the material either online or in the old-fashioned
dead-tree copies. Whether a university allows you to use their facilities or not is up to them.
- If you want to encourage public access to journals, you may want to look through this guide as to how you can help.
The author of that page is not going to be able to give you copies of the article you want, though, so don't
pester him!
Can I process my collection of ten thousand stories through this site?
If you do, you're liable to cause the site to crash and me to block your IP in a fit of annoyance. Email
me instead.
What software do you use?
Geeky namecheck coming up:
Ubuntu Linux,
Perl,
XML::Twig,
Lighttpd,
Lucene,
Solr,
LingMed,
and OpenCalais.
The site isn't identifying all the papers very well
Yes, this is sometimes the case, but I'm working on improving it as much as I can! In some cases the papers that
it returns can be a bit bizarre - for the moment, just vote these down by clicking the cross button, and help correct the features in the news text
on the left.
What are these "features", and how do I add and remove them?
Features are simply the names of people, places, organizations, and journals that are spotted in the news text. They appear in coloured
highlights. If you click on a highlighted word, a box should pop up (assuming you have javascript enabled) with an explanation of what it is, and
with an option to remove it if it's erroneous.
If you see the name of an author or journal that's not been identified already, then highlight it - this should popup a menu asking if you want to
add it as a feature (which will cause new papers to be searched for). For people, if someone's named as Professor Jane Smith, please only add "Jane
Smith", not her title as well!
If the correct paper's already been identified and voted for, these facilities won't be available.
I've got a great idea for your site / I've spotted a bug / I have another question / I'd like you to get your
site to work for my favourite news source / I work for an online news source and I'd like to talk
Email
me and let me know - I'm delighted to have suggestions, criticism and feedback. If you're reporting a bug, see
this page for advice to help me deal with it
better. This site is written and maintained in my spare time, so please be forgiving if I take a while to get back to
you.
What's the text behind the logo?
The Increase of the Badger (Meles meles) in Great Britain During the Period 1900-1934: With Especial Reference to an Area in South
Shropshire
Known as the Wheatland Country
Frances Pitt, The Journal of Animal Ecology, Vol. 4, No. 1 (May, 1935), pp. 1-6
Who are you?
I'm Adam Bernard, I live in Cambridge (the UK one). My day job is doing a PhD in text mining at the European Bioinformatics Institute. If you're interested in the EBI's literature
services, look here. thesciencebehindit.net is written entirely in my personal capacity. I like
lute music, fried-egg-sandwiches, and the Private Eye crossword.