Has anyone successfully integrated the AGROVOC Thesaurus (Agricultural Vocabulary) with GeoNetwork 2.1? Ideally I would like a search for “Corn” in my GeoNetwork keyword search and return all records containing the keyword “Maize” (for example). My understanding was that AGROVOC allows this type of synonymous (as well as broader / narrower) search. Unfortunately I’ve been getting a “null pointer exception” error every time I try and load the Thesaurus. Any thoughts?
Grant McKenzie GIS Developer
Enterprise Management Solutions
I have already try to integrate the Agrovoc thesaurus.
But the rdf file is more than 59Mo. If i remember well that cause an
exception because the file is too big.
Make a check in your log.
Anyway the release 2.1 doesn't manage the narrower and broader word. In the
2.1 the thesaurus is a simple list of keyword.
I don't know if the 2.2 manage this part.
Best regards,
Fabien Bachraty
GMcK wrote:
Has anyone successfully integrated the AGROVOC Thesaurus (Agricultural
Vocabulary) with GeoNetwork 2.1? Ideally I would like a search for
"Corn" in my GeoNetwork keyword search and return all records containing
the keyword "Maize" (for example). My understanding was that AGROVOC
allows this type of synonymous (as well as broader / narrower) search.
Unfortunately I've been getting a "null pointer exception" error every
time I try and load the Thesaurus. Any thoughts?
Grant McKenzie
GIS Developer
Enterprise Management Solutions
CH2M Hill
Seattle WA
Direct: 425.233.3271
Mobile: 425.233.4562
grant.mckenzie@anonymised.com
I have already try to integrate the Agrovoc thesaurus.
But the rdf file is more than 59Mo. If i remember well that cause an
exception because the file is too big.
Maybe try to not load the multilingual rdf file. Works fine for me even with the
big one. Problem with huge thesaurus, it's that the rdf file is loaded at
startup, if you use too big thesaurus that could slow down the GeoNetwork
startup process. So if you're using Agrovoc in only one language maybe filter
the language you want before loading.
Anyway the release 2.1 doesn't manage the narrower and broader word. In the
2.1 the thesaurus is a simple list of keyword.
I don't know if the 2.2 manage this part.
No it will not. An option could be to change the lucene index process to add
narrower, broader, related term to the index when indexing metadata, another
option is to add those terms at search time ...
I have already try to integrate the Agrovoc thesaurus.
But the rdf file is more than 59Mo. If i remember well that cause an
exception because the file is too big.
Maybe try to not load the multilingual rdf file. Works fine for me even
with the
big one. Problem with huge thesaurus, it's that the rdf file is loaded at
startup, if you use too big thesaurus that could slow down the GeoNetwork
startup process. So if you're using Agrovoc in only one language maybe
filter
the language you want before loading.
Anyway the release 2.1 doesn't manage the narrower and broader word. In
the
2.1 the thesaurus is a simple list of keyword.
I don't know if the 2.2 manage this part.
No it will not. An option could be to change the lucene index process to
add
narrower, broader, related term to the index when indexing metadata,
another
option is to add those terms at search time ...
Hi list, another issue which could occur with the Agrovoc thesaurus import is
in the latest release of Agrovoc (Date : 080422).
There is two non UTF-8 characters.
Here is line number in the SKOS file format.
diff ag_skos_080422.rdf
ag_skos_080422ok.rdf
Line number 29860 d 29859
< <skos:altLabel xml:lang="fa">مواد آرسنيك دار</skos:altLabel>
Line number 450448 d 450446
< <skos:altLabel xml:lang="fa">فوزاريوم اكسي اسپوروم
فازئولي</skos:altLabel>
Remove these lines (if you don't need it) to load the file without
exception.