Hi Thomas,

We haven't found an open thesaurus for engineering that works yet - Inspec has one, but it's subscription. The PLoS thesaurus is open and covers engineering, but it's not very deep because they don't publish a great deal in that area.

We want results for every paper because concept tagging provides inputs to processes that we perform on all papers (finding editors, finding reviewers, related papers, search, etc.) We don't want these processes to work better for some papers than for others.

We do use MeSH everywhere we can. When we can't, one solution is just to build a new scheme and publish it, as PLoS has done, integrating MeSH concepts into a larger vocabulary. This isn't so much a problem for us, it's just inefficient: all publishers are working on this problem now and building parallel but not exactly equivalent vocabularies. 

This is less elegant than if we used the same open vocabulary for all of science (way easier said than done). I'm not a taxonomist, but I think it leads to the need for a messy ontology layer to equate concepts from one publisher's corpus to another.

Andrew



On Sat, Nov 12, 2016 at 1:36 AM, Thomas Krichel <krichel@openlib.org> wrote:
  Andrew Smeall writes

> We do use MeSH for those subjects, but this only applies to about 40% of
> our papers. In Engineering, for example, we've had more trouble finding an
> open taxonomy with the same level of depth as MeSH.

  Have you found one?

> For most internal applications, we need 100% coverage of all
> subjects.

  Meaning you want to have a scheme that provides at least
  one class for any of the papers that you publish? Why?

> The temptation to build a new vocabulary is strong, because it's the
> fastest way to get to something that is non-proprietary and universal. We
> can merge existing open vocabularies like MeSH and PLOS to get most of the
> way there, but we then need to extend that with concepts from our corpus.

  I am not sure I follow this. Surely, if you don't have categories
  for engineering, you can build your own scheme and publish it. I don't
  see this as a reason for not using MESH when that is valid for the
  paper under consideration.

  I must be missing something.

--

  Cheers,

  Thomas Krichel                  http://openlib.org/home/krichel
                                              skype:thomaskrichel



--
------------------------------
Andrew Smeall
Head of Strategic Projects

Hindawi Publishing Corporation
Kirkman House
12-14 Whitfield Street, 3rd Floor
London, W1T 2RF
United Kingdom
------------------------------