search

shuyo / language-detection

This is a language detection library implemented in plain Java. (aliases: language identification, language guessing)
https://github.com/shuyo/language-detection/blob/wiki/ProjectHome.md
727 stars 184 forks source link

Unexpected language detected #96

Open drumnaoto opened 5 years ago

drumnaoto commented 5 years ago

Hello. Thank you for your this library, It was very helpful.

Well, I tried to detect one sentence... I expected the detected result is almost "en" , but "fr" is so high score.

Why did this result?? I hope you will answer. Thank you.

if you have any questions just let me know.
fr : 0.5714287861917584
en : 0.4285712109411652

If you have any questions just let me know
en : 0.5714271826718228
fr : 0.42857147940079865

If you have any questions just let me know.
fr : 0.8571390172993032)
en : 0.14286097221805769

if you have any questions just let me know.
en : 0.5714296257779372
fr : 0.4285694470833237
TobiasJu commented 5 years ago

Can confirm, just skipping the dot makes the sentence English. With dot at the end it's french. But why?