Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Mixed languages polarized by "en" #60

Open
loretoparisi opened this issue Oct 19, 2016 · 1 comment
Open

Mixed languages polarized by "en" #60

loretoparisi opened this issue Oct 19, 2016 · 1 comment

Comments

@loretoparisi
Copy link

I have the following text

that is a mix of english and sesotho:

>>>Ska rebona re phela\nKgale re sokola rona re phelela mmino\nO skang potja ka dilo\nKgale re sokola rona re phelela mmino (×2)\nWe five minutes from freedom\nSomebody tell my mama

Having the whole sentence as-it-is I get a wrong identification:

>>> Ska rebona re phela\nKgale re sokola rona re phelela mmino\nO skang potja ka dilo\nKgale re sokola rona re phelela mmino (×2)\nWe five minutes from freedom\nSomebody tell my mama
('en', -233.38300132751465)

Removing the whole en sentences, and I get the right ISO-639-1 language code: sl:

>>> Ska rebona re phela\nKgale re sokola rona re phelela mmino\nO skang potja ka dilo\nKgale re sokola rona re phelela mmino                                                       
('sl', -154.34662437438965)

Also keeping only one en sentence, the right language is recognized:

>>> Ska rebona re phela\nKgale re sokola rona re phelela mmino\nO skang potja ka dilo\nKgale re sokola rona re phelela mmino (×2)\nWe five minutes from freedom                    
('sl', -217.8226833343506)

So, it seems that the detector is being "polarized" by the en sentences in this phrase.

@sixtyfive
Copy link

One would guess that this is due to bias from training.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants