5 ms·

I wish it just struggled with the mix of multiple languages, it can't even handle my native one. A plural compound word in translative case with a clitic? I can

I wish it just struggled with the mix of multiple languages, it can't even handle my native one. A plural compound word in translative case with a clitic? I can only dream about it, it can't even do all the cases reliably.

Aachen 6y ago
What language is that? I guess what I'm wondering is how obscure it has to get (or perhaps how low the GDP of the people speaking a language has to be) before custom spell checking rules aren't considered worth it to bigcorps anymore. Though I'm also kinda interested in what this weird language thing is you're talking about.
- hoseja 6y ago
  Based on the features; Hungarian, Finnish, or Estonian.
  - vesinisa 6y ago
    Microsoft has an okay Finnish spellchecker. It's really a shame no Finnish university has developed an acceptable open source Finnish spellchecker though AFAIK. Word-list based checkers actually work for English, which gives an impression that they would work for all types of languages. For agglutinative languages (which I think is about 50% of world languages) not so much, since the list of valid words is basically infinite. But combined with a high-quality root word list it's possible to write a computer program that returns if a given input word is a valid form of some known root word, i.e. a spellchecker. In English, you can think the problem same as writing a spellchecker that can tell if an input word is a valid chemical name for a molecule, like 1,3,7-trimethylxanthine (caffeine) or adenosine-5'-triphosphate (ATP). Clearly, we can not write an exhaustive list of valid words here. But it's pretty easy to write a computer program that can tell if the input word at least seems like a valid chemical name devoid of typos.
- afiori 6y ago
  Could be Italian, where there are words like `portatemelo` ~ `you bring that to me` that google routinely gets wrong.

Aachen 6y ago

What language is that? I guess what I'm wondering is how obscure it has to get (or perhaps how low the GDP of the people speaking a language has to be) before custom spell checking rules aren't considered worth it to bigcorps anymore. Though I'm also kinda interested in what this weird language thing is you're talking about.

hoseja 6y ago
Based on the features; Hungarian, Finnish, or Estonian.
- vesinisa 6y ago
  Microsoft has an okay Finnish spellchecker. It's really a shame no Finnish university has developed an acceptable open source Finnish spellchecker though AFAIK. Word-list based checkers actually work for English, which gives an impression that they would work for all types of languages. For agglutinative languages (which I think is about 50% of world languages) not so much, since the list of valid words is basically infinite. But combined with a high-quality root word list it's possible to write a computer program that returns if a given input word is a valid form of some known root word, i.e. a spellchecker. In English, you can think the problem same as writing a spellchecker that can tell if an input word is a valid chemical name for a molecule, like 1,3,7-trimethylxanthine (caffeine) or adenosine-5'-triphosphate (ATP). Clearly, we can not write an exhaustive list of valid words here. But it's pretty easy to write a computer program that can tell if the input word at least seems like a valid chemical name devoid of typos.
afiori 6y ago
Could be Italian, where there are words like `portatemelo` ~ `you bring that to me` that google routinely gets wrong.