Exporting a selected ConceptScheme of a Project in version 10.0.1

[Jeannine Beeken]

Hi,

I have two schemes in a project, one of which contains a partition of the concepts of the other tree (created using the option 'add selected concept's subtree to scheme...).

There is a question/answer re exporting only 1 scheme of a project, dating form 2018 at https://groups.google.com/g/vocbench-user/c/l_V7YuPiNfI/m/878ULNLkAgAJ, where you provided a Sparql query for the Export option as it was defined then and the steps/process 'Global Data Management -> Export data -> Export filter -> SParqlExportFilterFactory.

However, the export facility of the latest VB version and its available options have been thoroughly revised and seem to be completely different from earlier versions.

Would it be possible to both describe the different steps of the export facility/process in version 10, and indicate whether the Sparql query mentioned in the link given above is still valid, esp. where/how to add it ' Global Data Management -> Export data > Data transformations -> ? -> ?

Thank you and best wishes,

Jeannine 

[Tiziano Lorenzetti]

Dear Jeannine,

sorry for the late reply. The solution proposed in the thread you linked is still valid, we just changed some terminology in the UI:

• Export filter -> Data transformations
• SPARQLExportFilterFactory -> SPARQL RDF transformer
  

Apart from these, the described procedure has remained unchanged.

Best regards,

Tiziano

--
You received this message because you are subscribed to the Google Groups "vocbench-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vocbench-use...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/vocbench-user/5fc714d7-e61d-4cf5-a403-66c8b6f31aban%40googlegroups.com.

[Jeannine Beeken]

Hi,

I tried again to export one of the schemes in the latest v 10, but without success. I followed the instructions below and tried for

1) Deployment: save to file - Reformatter: Spreadsheet serializing exporter - Export format: XLSX - Include inferred not ticked

2) Deployment: save to file - Reformatter: RDF serializing exporter - Export format: RDF/XML - Include inferred not ticked

and submitted the queries. The system started loading each time, but never came back with a result.  Could you please advise what to do, thanks.

Best wishes,

Jeannine

[Roland Wingerter]

Dear Jeannine,

I exported a selected concept scheme from the "Land and Water" thesaurus (see VocBench documentation) and it worked for me. You said the system "never came back with a result". Note that processing the SPARQL query will take some time when you have a large thesaurus (I first tried using the same procedure to export data from Eurovoc, which has 12 million triples, but did not have the patience to wait until the procedure was finished.)

Kind regards

Roland

[Armando Stellato]

Hi everybody,

 

one other possibility is that the download is ready but your browser is preventing it. Have you checked if there’s any popup/download blocker active?

 

Kind Regards,

 

Armando

 

P.S: if your data is small and doesn’t contain private information you can send it directly to me. I’ll give it a try

 

--

You received this message because you are subscribed to the Google Groups "vocbench-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vocbench-use...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/vocbench-user/bd976f29-d255-4736-b69e-95d6c9ab041bn%40googlegroups.com.

[Beeken, Jeannine C T]

Hi Armando,

 

Thank you for your suggestion. I am using Firefox;  I unticked the popup blocker and submitted the query a minute ago. Before doing that, I intentionally added a blank at two spots of the query, which the programme detected as errors, and removed them again, so that works. The programmes started loading and I am waiting now…

 

Best wishes,

Jeannine

 

From: Armando Stellato <sta...@outlook.it> On Behalf Of Armando Stellato
Sent: 31 May 2022 08:18
To: Roland Wingerter <chun...@gmail.com>; vocbench-user <vocben...@googlegroups.com>
Subject: RE: [vocbench-user] Re: Exporting a selected ConceptScheme of a Project in version 10.0.1

 

CAUTION: This email was sent from outside the University of Essex. Please do not click any links or open any attachments unless you recognise and trust the sender. If you are unsure whether the content of the email is safe or have any other queries, please contact the IT Helpdesk.

To view this discussion on the web visit https://groups.google.com/d/msgid/vocbench-user/AS8PR09MB4982334BA62D1E972C37CFA3C7DC9%40AS8PR09MB4982.eurprd09.prod.outlook.com.

[Beeken, Jeannine C T]

Hi Armando,

 

An update: the programme started loading at 9am and is still loading at nearly 1pm (UK time).  It concerns an ELSST thesaurus project which contains two Schemes: the whole thesaurus (in 10+ languages having prefLabels, altLabels and many notes) and a small multilingual selection, namely all resources from one topTerm down to its lowest NT level (using: a small tree of 20x10+ resources. 

Would it be an option to have a query which is not negative, i.e. deleting/filtering out all concepts that do not appear in the scheme to be exported, but rather positive, i.e. selecting/filtering out the concepts of the scheme only without any comparison, thanks. Since the Loading is slowing down all other apps and actions, and ‘freezing’ the VB app, I refreshed it without logging out.

The following Error message appears twice, when trying to go to the project Data: org.eclipse.rdf4j.query.QueryEvaluationException: java.io.EOFException

However, when trying again, it seemed to work and the data/concepts are displayed again. I also created another scheme/subtree now to check whether the first one contained the same metadata and values, and it did, indeed.

 

Best wishes,

Jeannine

[Andrea Turbati]

Dear Jeannine,

did manage to export the desired ConceptScheme at the end? If not, let me know if I can help in any way (for example, if the data is not too big and does not contain any private information you could sent directly to me so I can have a look and try to understand what the problem is ). Exporting the data "belonging" to a single ConceptScheme could be, depending on the data itself, not an easy task not just technically but also theoretically, since it is important to decide which data to export when complex relationships are present in the data itself.

Kind Regards,

Andrea

To view this discussion on the web visit https://groups.google.com/d/msgid/vocbench-user/DB7PR06MB44577A05970A5951353E9BDBBEDC9%40DB7PR06MB4457.eurprd06.prod.outlook.com.

[Beeken, Jeannine C T]

Dear Andrea,

 

Thank you for your mail.  As it happens, I tried again last week. This time I did not get a time-out, but after waiting 8 hours for the ‘Uploading’ to end, I logged out. No export was delivered (I asked for an export as file/Excel sheet, no inferences to be added)

It concerns the ELSST thesaurus as such, having created two small subsets/schemes for export (see my mail below). I selected one of the subsets (completed the Sparql example with its URI, but, as said, it might be an issue of size; there is no private information involved. As suggested in my mail below,” Would it be an option to have a query which is not negative, i.e. deleting/filtering out all concepts that do not appear in the scheme to be exported, but rather positive, i.e. selecting/filtering out the concepts of the scheme only without any comparison, thanks.”

 

Best wishes,

Jeannine

To view this discussion on the web visit https://groups.google.com/d/msgid/vocbench-user/6561166e-568a-301d-c3a1-587d46698bb6%40gmail.com.

[Andrea Turbati]

Dear Jeannine,

in VocBench when doing an export two possible scenarios are possible:

1. the user does NOT specify any "data transformations" element, so the data is exported as it is (only the selected graphs)
2. the user specify at least one "data transformations" and in this case a temporary copy of the data is copied in memory and all the desired transformations are done on this in-memory copy (so the original data is not modify in any way), so, the transformations are mostly done by removing some RDF triples (but you can also add/transform some of them)

Since you say that the data is quite big, you could try to provide more memory to SemanticTurkey (the server behind VocBench) and this could help.

Another solution in this case (if you are having a problem with the size/complexity of the data) could be to do a complete export of the project, load this export in a NEW project in VocBench and then "manually" (using SPARQL updates for examples) remove the data from the other scheme(s) and, once the new project contains only the desired data, you could then do a complete export of this new project. Naturally, if the data from the original project is changed, then you need to to this whole process again (so this is solution which should be used only then the standard export procedure, with the data transformers, does not word due to the size or the complexity of the original data).

Kind Regards,

Andrea