robovoice
(robovoice)
January 16, 2022, 12:00pm
1
#3378
I filed a request on github for this.
What is the opinion on this?
opened 08:58AM - 03 Dec 21 UTC
closed 05:44PM - 21 Jan 24 UTC
Discussion
The idea of this request is:
The contributer, who is not submitting (useful!)… clips for the (selected) languages and/or dialects cannot validate for that language and/or dialect. (based on the displayed languages in CV settings profile and stats, your languages, see pic1)
Not selected languages in CV website cannot be reached for a speak/validating session. (see pic2)
Submitting useful clips is the proof of speaking and understanding the selected language(s).
Some other things to think of:
Are these limitations for newbies only or permanent?
Has the non native speaker the right to judge on submitted native speaker clips?
Is this request improving the validation process/results of validation or just adding new obstacles to the contributor?
This request could prevent "messing around" in language sections, which the contributor is not able to speak or understand.
**Additional context**
Add any other context or screenshots about the feature request here.


bozden
(Bülent Özden)
January 16, 2022, 2:29pm
2
opened 08:58AM - 03 Dec 21 UTC
closed 05:44PM - 21 Jan 24 UTC
Discussion
The idea of this request is:
The contributer, who is not submitting (useful!)… clips for the (selected) languages and/or dialects cannot validate for that language and/or dialect. (based on the displayed languages in CV settings profile and stats, your languages, see pic1)
Not selected languages in CV website cannot be reached for a speak/validating session. (see pic2)
Submitting useful clips is the proof of speaking and understanding the selected language(s).
Some other things to think of:
Are these limitations for newbies only or permanent?
Has the non native speaker the right to judge on submitted native speaker clips?
Is this request improving the validation process/results of validation or just adding new obstacles to the contributor?
This request could prevent "messing around" in language sections, which the contributor is not able to speak or understand.
**Additional context**
Add any other context or screenshots about the feature request here.


I fail to understand the rationale behind this suggestion. Did you encounter many spammers who record in languages completely unknown to them?
For example, my first foreign language was German, but I’m a bit (lot) rusty. Whenever I have time I want to add German and start to validate to get rid of this rust. Afterward, I can record.
In case of foreign language speakers, CV wants to record their voices.
2 Likes
robovoice
(robovoice)
January 16, 2022, 5:23pm
3
Link was broken, thanks for re-posting!
After some time recording the contributor may starts validating sentences.
The proposal was for newbies, non native speakers with very basic skills, and yes, trolling could also be a reason.
The " rationale" could be: A selected time perpiod in which the contributor can prove, that his results (clip contributing and validation) are useful for the corpus also affecting later training the model(s).
1 Like
bozden
(Bülent Özden)
January 16, 2022, 5:36pm
4
Two major deal-breakers for this or similar feature:
The CV database does not keep statistics about user behavior.
You can even contribute without registering.
Similar changes were discussed here:
What does Community Health mean to you ?
As part of the Common Voice Community strategy, we are thinking of ways to support sustainable and healthy communities across common voices.
One of the ideas, is the development of community health metrics, to help us have a wide picture of the health of language communities. In turn, helping communities self-organise effectively and for me prioritise support across our growing community.
Data collection and metrics
I have taken inspiration from CHAOSS…
internetman
(terance@protonmail.com)
June 6, 2022, 11:29pm
5
Could a possible midlde ground for this be that CV keeps track of how big percentage of validations from a user is also validated (either yes/no) by other users who has alot of recorded clips?