By the way, the script does not generate the true top level, which is the leading letter:

 

·         Anatomy [A]

·         Organisms [B]

·         Diseases [C]

·         Chemicals and Drugs [D]

·         Analytical, Diagnostic and Therapeutic Techniques and Equipment [E]

·         Psychiatry and Psychology [F]

·         Biological Sciences [G]

·         Physical Sciences [H]

·         Anthropology, Education, Sociology and Social Phenomena [I]

·         Technology and Food and Beverages [J]

·         Humanities [K]

·         Information Science [L]

·         Persons [M]

·         Health Care [N]

·         Publication Characteristics [V]

·         Geographic Locations [Z]

 

This is left as an exercise for the reader, as they say.

 

Andy

 

 

 

OK, here it is:

MESH Subjects file

Perl Script

MESH source file

MESH toplevel headings

 

http://researchonline.lshtm.ac.uk/2965187/

 

Enjoy

 

Andy

 

 

I’m looking at a similarly large source – sounds like it would be useful to see your code if only to avoid obvious pitfalls.

 

I suppose the ideal would be to have a tool that could handle a range of different vocabularies – Mesh seems like an obvious candidate for testing as it’s so comprehensive.

 

Lizz

 

 

 

I did try converting the MESH (MEdical Subject Headings) ontology into an Eprints Subject tree, back when we started about five years ago.  I succeeded insofar as the script built a subject file that could be imported correctly in principle.  But it was so vast that the test server ground to a halt – I think there were 50000+ terms.  Also, the problem is that MESH is not really one ontology, but about twenty separate dimensions: Disease, Body Part, Geography, Study Population, etc etc.  I can probably dig out the script and/or the load file if anyone wants to have a play with it.  Probably only of interest to Medical Schools.

 

Andy

 

 

Andy Reid

Research Information Manager

Executive Office, Room G40a

London School of Hygiene and Tropical Medicine

Keppel St, LONDON, WC1E 7HT

0207-927-2618 (Internal/Teleworker x2618)

 

 

 

 

Hello all,

 

Has anyone implemented a thesaurus in place of the standard subject tree in EPrints?

 

Has anyone tried to implement a thesaurus in place of the standard subject tree in EPrints and found it impossible because of some fundamental incompatibility?

 

I would be interested in hearing any experiences before planning work in this area!

 

Lizz

 

--

Lizz Jennings BA MSc ACLIP MCLIP (Revalidated 2015)

Research Data Librarian (Systems)

The Library 4.10, University of Bath, Bath, BA2 7AY UK

Ext. 3570 (External 01225 383570)

E.Jennings@bath.ac.uk

Research Data Management: http://www.bath.ac.uk/research/data