ELTeC Core / ELTeC extension

[CarolinOdebrecht]

This proposal in no way invalidates the sampling criteria. In our action we will have as a core output a corpus created according to uniform criteria. In the following I will make a distinction between ELTeC Core (all sampling and balancing criteria are met) and an ELTeC extension.

To give visibility to the different publishing traditions and to find a practical way to deal with text material that cannot be integrated into ELTeC Core, we can create a second extension collection - ELTeC Extension. This would allow Language Collections, which have already integrated the 100 necessary texts that meet the criteria, to add more texts. Again, languages that cannot quite reach the 100 texts (with all criteria) can also add further texts (e.g. from authors who are already sufficiently represented in ELTeC core) to this ELTeC extension.

If we want to follow this, we still have to talk about the workflow.
Should the texts of ELTeC Extension be placed in a separate folder in the ELTeC core language repositories?
Should we set up our own organization for this?
I would recommend the latter.

[dianamsmpsantos]

Dear Caroline Before you decide your own organization or whatever that might me, it would be very practical for us Portuguese, who have put more books than the ones allowed from the same author, to be able to put them asap in a new folder called extension, under por-eltec. If you could provide us with such a folder, we would be very grateful. Diana CarolinOdebrecht <notifications@github.com> escreveu no dia segunda, 28/01/2019 à(s) 09:24:

…

This proposal in no way invalidates the sampling criteria. In our action we will have as a core output a corpus created according to uniform criteria. In the following I will make a distinction between ELTeC Core (all sampling and balancing criteria are met) and an ELTeC extension. To give visibility to the different publishing traditions and to find a practical way to deal with text material that cannot be integrated into ELTeC Core, we can create a second extension collection - ELTeC Extension. This would allow Language Collections, which have already integrated the 100 necessary texts that meet the criteria, to add more texts. Again, languages that cannot quite reach the 100 texts (with all criteria) can also add further texts (e.g. from authors who are already sufficiently represented in ELTeC core) to this ELTeC extension. If we want to follow this, we still have to talk about the workflow. Should the texts of ELTeC Extension be placed in a separate folder in the ELTeC core language repositories? Should we set up our own organization for this? I would recommend the latter. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#23>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AHx-csFuTNpGAx6mUh4eclAkuMGO1twyks5vHrOogaJpZM4aVZ3D> .

[lb42]

Diana's idea sounds good to me. Let's just have an additional folder in each affected repo called "Overflow" or "Reserve" or similar, and move surplus texts there. No need for any special effort. (Diana -- you can create such a folder yourself). This also makes it much easier to fine tune the balance of the corpus as additional texts become available. (Obviously it doesnt help at all in the case where there are insufficient texts for whatever reason)

[CarolinOdebrecht]

Dear Diana and Lou, dear all, I agree that to add additonal folder(s) is a pragmatic way to handle this. I would suggest to have a "extensionlevel0" and "extensionlevel1" folder in order to sort the files by their encoding level. We could then consider later whether an own repo or organization is really necessary.