female forms and ili

[hdaSprachtechnologie]

The new DTD allows to have a synset relation "has female". This could be used, for example, for entries "Arzt" and "Ärztin", the male and female form of "doctor". Currently, both belong to the same synset, which is wrong. But, on the other hand, if we have two synsets and a relation between them, what should be done with the ili? We would again introduce cases where two synsets show to the same ili. Or is there another solution to this problem?

[fcbond]

Hi, I think there are two possible approaches. If you think the 'male' form is unmarked (so you use it when gender is unknown) then linking that to the ili (which is linked to English which is unmarked) and having "Ärztin" linked to it with no ili (or possibly request a new ili) seems best. Alternatively, you could have an unlexicalized synset 'neutral doctor' with children "Arzt" and "Ärztin". If I remember correctly, Polish Wordnet does the first and Bulgarian the second. I think for odenet maybe the first makes most sense, what do you think?

…

On Mon, Dec 20, 2021 at 11:54 PM Melanie Siegel ***@***.***> wrote: The new DTD allows to have a synset relation "has female". This could be used, for example, for entries "Arzt" and "Ärztin", the male and female form of "doctor". Currently, both belong to the same synset, which is wrong. But, on the other hand, if we have two synsets and a relation between them, what should be done with the ili? We would again introduce cases where two synsets show to the same ili. Or is there another solution to this problem? — Reply to this email directly, view it on GitHub <#41>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAIPZRRGNR5RDZWPPZAOIWDUR5GSDANCNFSM5KOA6ZEA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

-- Francis Bond <https://fcbond.github.io/> Division of Linguistics and Multilingual Studies Nanyang Technological University

[hdaSprachtechnologie]

Thank you!
I have now added "Ärztin" with a new synset and empty ili and a has_female/has_maskuline relation. Will work on other female forms in the next time.

[fcbond]

Great, it is nice to see the tags used. Odenet may be the first! I think in this case, it is relevant to tag both sense_relation (as it is derived) and synset_relation (as the concepts are meaningfully linked).

…

On Tue, Dec 21, 2021 at 11:07 PM Melanie Siegel ***@***.***> wrote: Thank you! I have now added "Ärztin" with a new synset and empty ili and a has_female/has_maskuline relation. Will work on other female forms in the next time. — Reply to this email directly, view it on GitHub <#41 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAIPZRQA5B6KHINBFEVRTO3USCJZNANCNFSM5KOA6ZEA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you commented.Message ID: ***@***.***>

-- Francis Bond <https://fcbond.github.io/> Division of Linguistics and Multilingual Studies Nanyang Technological University

[arademaker]

Currently, both belong to the same synset, which is wrong.

Why? I guess the definition and relations must support the general meaning and not the specific masculine/feminine ones. I am not saying I don't agree with the ideia, in the Portuguese https://github.com/LR-POR/MorphoBr we also take the feminine and masculine nouns as two distinct words. That is, nouns inherently carry one value of gender.

But taking that view in OWN-PT will be a huge change... I would be interested the arguments to support it.

[fcbond]

Hi, I think the masculine and feminine forms clearly have different meanings, and are *not* typically substitutable for one another, which is the general guide for what goes into the same synset/synonymy: so that is the motivation to split them. But they do have clearly related meanings and this relation is slightly different from normally hypernymy, which is why we created the new relations.

…

On Thu, Dec 23, 2021 at 9:45 PM Alexandre Rademaker < ***@***.***> wrote: Currently, both belong to the same synset, which is wrong. Why? I guess the definition and relations must support the general meaning and not the specific masculine/feminine ones. I am not saying I don't agree with the ideia, in the Portuguese https://github.com/LR-POR/MorphoBr we also take the feminine and masculine nouns as two distinct words. That is, nouns inherently carry one value of gender. But taking that view in OWN-PT will be a huge change... I would be interested the arguments to support it. — Reply to this email directly, view it on GitHub <#41 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAIPZRXNYFNBOZ4I25QQWF3USMRXNANCNFSM5KOA6ZEA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you commented.Message ID: ***@***.***>

-- Francis Bond <https://fcbond.github.io/> Division of Linguistics and Multilingual Studies Nanyang Technological University