Lv1VillagerA
Lurker
- Joined
- Apr 3, 2018
- Messages
- 1,370
- Reputation score
- 3,082
Hi,
I've been recently working on an auto-translator and some game translations.
I'm facing a somewhat challenging problem: the translation of onomatopoeia/interjections (I've narrowed down the problem to listing and categorizing them in english).
I've seen many different ways of translating those ranging from creative to lazy; none have convinced me.
Machine translation, and google translation more specifically, can't translate those because most of the syntax is extremely specific or invented by the author (no way a learning algorithm would learn without being fed lots of hentai): no trivial solution.
I'm writing an auto-translator so I need a reliable and methodical way to translate those and I know there is at least one.
After spending some time manually translating those, I've basically gave up on the idea a simple algorithm will ever translate JAP->ENG. This is why I came up with a new approach, with the algorithm detecting the mood of the word (very easy, う is a negative mood, は is an exclamation...) and only then re-writing it to english.
(I've already solved half of my problem by writing a simple algorithm to isolate onomatopoeia/interjections.)
Now I need a large and structured english corpus of English onomatopoeia/interjections. For the structure of the corpus, I had the idea of something like
Then I'll be able to have:
I there's any english hentai writer (or enthusiast) out there, feel free to post you best moans, cries of terror, etc...
Maybe you too are interested in having more expressive protagonists and sex scenes.
I've been recently working on an auto-translator and some game translations.
I'm facing a somewhat challenging problem: the translation of onomatopoeia/interjections (I've narrowed down the problem to listing and categorizing them in english).
Ahhnn
Tsk.
Hmph
Phiew
Huff
Wha...?
No!
Tsk.
Hmph
Phiew
Huff
Wha...?
No!
はぁぁぁぁぁぁぁ
はぁぁぁぁぁぁぁぁ
はぁぁぁぁぁぁ
うっはぁぁぁぁぁ
んはぁぁぁぁぁ
ふわぁぁぁぁぁ
ふわぁぁぁっ！
あぁぁっ！
うらぁっ！
はぁぁぁぁぁぁぁぁ
はぁぁぁぁぁぁ
うっはぁぁぁぁぁ
んはぁぁぁぁぁ
ふわぁぁぁぁぁ
ふわぁぁぁっ！
あぁぁっ！
うらぁっ！
I've seen many different ways of translating those ranging from creative to lazy; none have convinced me.
Machine translation, and google translation more specifically, can't translate those because most of the syntax is extremely specific or invented by the author (no way a learning algorithm would learn without being fed lots of hentai): no trivial solution.
I'm writing an auto-translator so I need a reliable and methodical way to translate those and I know there is at least one.
After spending some time manually translating those, I've basically gave up on the idea a simple algorithm will ever translate JAP->ENG. This is why I came up with a new approach, with the algorithm detecting the mood of the word (very easy, う is a negative mood, は is an exclamation...) and only then re-writing it to english.
(I've already solved half of my problem by writing a simple algorithm to isolate onomatopoeia/interjections.)
Now I need a large and structured english corpus of English onomatopoeia/interjections. For the structure of the corpus, I had the idea of something like
You must be registered to see the links
.Then I'll be able to have:
Surprised, Confused (Eng): Wha...; How?, Really?; etc...
Surprised, Confused (Jap): includes:"・・・", "え", "！", etc.
Surprised, Confused (Jap): includes:"・・・", "え", "！", etc.
I there's any english hentai writer (or enthusiast) out there, feel free to post you best moans, cries of terror, etc...
Maybe you too are interested in having more expressive protagonists and sex scenes.