PLoS One, Text-mining, Metrics and Bats

Just heard that PLoS One was awarded Innovator of the Year by SPARC:

http://blogs.plos.org/everyone/2011/06/30/plos-one-wins-recognition-as-a-sparc-innovator/

I applaud them personally as the 4 Pantonistas were given the same award last year for the Panton Principles.

So Lezan, collaborators at NaCTEM and I have published our first article in PLoS:

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0020181

Using Workflows to Explore and Optimise Named Entity Recognition for Chemistry

Top of Form

Bottom of Form

BalaKrishna Kolluru1*, Lezan Hawizy2, Peter Murray-Rust2, Junichi Tsujii1, Sophia Ananiadou1

For those who don’t know, PLoS One publishes competent science. Not trendy science, not stuff that the marketeers think will sell the journal. Just plain competent science. Simply:

  • We have said we have done X,Y,Z
  • It is in scope for the journal (a lof of science is out of scope for PLoS one
  • The referees agreed that we had done X,Y,Z competently. No “this isn’t interesting”, “not sufficient impact”.

To be honest it took an AWFUL long time to get it reviewed. SIX MONTHs (look at the dates) to get two referees opinions. I doubt this is specific to PLoS, it’s the fundamental problem of refereeing, to which there is no good answer.

Anyway it has been out for a few weeks? What does the world think of it? Well it has been out about 6 weeks and had 316 downloads. That’s’ exciting to young scientists. 300 people have clicked on their article. (Maybe they haven’t READ it, but at least it’s an indication). And another of Lezan’s papers has got a “Highly accessed” in J. Chem Informatics (http://www.jcheminf.com/content/3/1/17)

Accesses
475 Research article    
ChemicalTagger: A tool for semantic text-mining in chemistry
Lezan Hawizy, David M Jessop, Nico Adams, Peter Murray-Rust
Journal of Cheminformatics 2011, 3:17 (16 May 2011)
[Abstract] [Full Text] [PDF] [PubMed] [Related articles]

Well I am not a great fan of metrics of any sort. We have ahd 300,000 downloads of our software (official Microsoft figures) and we get zero credit. But at least we have a few hundred downloaders. So is 300 good? Impossible to say, but I’ll have a little fun with metrics:

Lets’ go to PLoS on May 27 and see the other article downloads. They’re 512, 322, 511, 295, 458, 493, 398 … So Lezan and Bala are within the range. Good, competent, science. (Text-mining science cannot be trendy because if we actually try to do it we’ll be sued by the published for mining “their” content – it is deeply depressing to be prevented from doing science by lawyers).

So what’s the sort of access for a highly accessed article? Go to http://www.plosone.org/home.action and “most viewed” and there are articles 1 week old with several thousand views. What the record? This one about bats:

http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0007595

It’s just under 2 years and had nearly 200,000 accesses. If it were a physical journal it would have fallen apart.

It’s had about 2 citations, which shows how stupid these metrics are

“download” it and see why it’s popular. You might even read it (I did, briefly)

But of course that will distort the metrics. Open access encourages people to READ articles. Whereas articles are actually only meant to be cited, not read.

 

This entry was posted in Uncategorized. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *