Vocbench index in non latin script

[Katerina Gkirtzou]

Dear Vocbench community,
I installed in a local machine the vocbench 8.0.1 and ontotext graph db version 9.1.1 using docker-compose. Within vocbench I create an OntoLex project with remote access to ontotext graph db and load my data from vocbench into the ontotext triple store.

The problem that I face is that all the written representations and the rdfs:label are in Greek, thus the index functionality you provide with a latin script, is not really useful to our case. Is there a way to declare or configure which script the index would be? The index functionality is really useful, as the lexicographers could have an overview of the entries quickly.

Furthermore, as our lexical resource contains approximately 69000 lexical entries, in order to start having something shown at the index, we need to search something first. In the initial search, the system warn us with the message below.

When we force the system to index, we do get only lexical entries that start with the first letter of our search term. But the system becomes slow in response. Do you think if we provide more memory or CPU (we are using the default ones) could assist to the responsiveness?

Any other tip/suggestion to help us?

Best regards,

Katerina Gkirtzou

[Armando Stellato]

Dear Katerina,

 

This has been fixed here:

 

https://bitbucket.org/art-uniroma2/semantic-turkey/commits/8cc6c9182e65e833c5cdae64228f46e028d571e5

 

which happens to be before version 9.0

 

Could you give version 9.0 (or the most recent 9.1) a try and let us know if you still have issues with the search?

 

Kind Regards,

Armando

 

 

From: vocben...@googlegroups.com <vocben...@googlegroups.com> On Behalf Of Katerina Gkirtzou
Sent: Monday, July 12, 2021 4:10 PM
To: vocbench-user <vocben...@googlegroups.com>
Subject: {Spam?} [vocbench-user] Vocbench index in non latin script

 

Dear Vocbench community,
I installed in a local machine the vocbench 8.0.1 and ontotext graph db version 9.1.1 using docker-compose. Within vocbench I create an OntoLex project with remote access to ontotext graph db and load my data from vocbench into the ontotext triple store.

The problem that I face is that all the written representations and the rdfs:label are in Greek, thus the index functionality you provide with a latin script, is not really useful to our case. Is there a way to declare or configure which script the index would be? The index functionality is really useful, as the lexicographers could have an overview of the entries quickly.

Furthermore, as our lexical resource contains approximately 69000 lexical entries, in order to start having something shown at the index, we need to search something first. In the initial search, the system warn us with the message below.

 

When we force the system to index, we do get only lexical entries that start with the first letter of our search term. But the system becomes slow in response. Do you think if we provide more memory or CPU (we are using the default ones) could assist to the responsiveness?

Any other tip/suggestion to help us?

 

Best regards,

Katerina Gkirtzou

--
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/38b6c9cd-85f8-4161-bd8a-1697cb38e9ffn%40googlegroups.com.