Question SYSTRANSOFT SYSTRAN v6 better than google translation

Discussion in 'Novel General' started by Deadmantellnotales, May 14, 2019.

  1. Deadmantellnotales

    Deadmantellnotales Member

    Joined:
    Dec 18, 2018
    Messages:
    18
    Likes Received:
    2
    Reading List:
    Link
    is the SYSTRANSOFT SYSTRAN v6 is better than google traslation. found the softwere in torrent.
    some of the series i been following are either dropped or didnt even translated.so whats tour opinion???
     
    Geek reader likes this.
  2. Westeller

    Westeller Happy kimono day~ Staff Member

    Joined:
    Apr 3, 2016
    Messages:
    5,872
    Likes Received:
    17,325
    Reading List:
    Link
    My opinion is that there are various options for machine translation and they all ultimately suck. You have two real choices if you actually give a damn: you can use Translation Aggregator to compare multiple machine translation results and piece the novel together like a puzzle, or you can use a glossary of known terminology to dramatically improve the results of any given machine translator.

    Or both together, I suppose.

    If either choice sounds like too much work to you, give up on getting readable results from MTL.
     
  3. listener

    listener ##################

    Joined:
    Apr 23, 2019
    Messages:
    1,148
    Likes Received:
    1,972
    Reading List:
    Link
    i thought it's only have a web version though....
     
  4. vancenick

    vancenick Well-Known Member

    Joined:
    Jun 23, 2016
    Messages:
    244
    Likes Received:
    139
    Reading List:
    Link
    INB4 keylogged and hacked
     
    Cream likes this.
  5. unkxz

    unkxz Lurker

    Joined:
    Nov 27, 2015
    Messages:
    398
    Likes Received:
    344
    Reading List:
    Link
    TBH, for me I think the most important thing of MTL is not how correct or accurate it is but how comfortable you're reading the translation. so find the translator you're comfortable with and go with it.
     
  6. AliceShiki

    AliceShiki 『Ms. Tree』『Ophi-kun's Survival Teacher』

    Joined:
    Apr 27, 2016
    Messages:
    17,056
    Likes Received:
    70,239
    Reading List:
    Link
    I tried using Systran once and it was honestly the worst translation software I ever touched (at least for Japanese Novels), Google and Bing were waaaaaaay better, and those 2 are pretty bad.
     
    checkm8 likes this.
  7. listener

    listener ##################

    Joined:
    Apr 23, 2019
    Messages:
    1,148
    Likes Received:
    1,972
    Reading List:
    Link
    and how about yahoo japan ?
     
  8. UDI

    UDI 『SCP-9369-EggMan』/ Lime Blood / Maid Of Space

    Joined:
    Jul 29, 2017
    Messages:
    305
    Likes Received:
    273
    Reading List:
    Link
    How would one actually go about doing this?
     
  9. Westeller

    Westeller Happy kimono day~ Staff Member

    Joined:
    Apr 3, 2016
    Messages:
    5,872
    Likes Received:
    17,325
    Reading List:
    Link
    A glossary of known terminology - sometimes also called a dictionary - is just a list of terms with manual translations. When the machine translator finds those terms, it uses the translation you've provided instead of trying to work it out on its own. This is especially useful to avoid the confusion that names cause, along with other common novel terminology. I'm sure you can imagine how much it improves the accuracy. Of course, for serious improvement you want to have as comprehensive a glossary as possible. LNMTL uses glossaries, and the top novels there have hundreds to thousands of pre-defined terms.

    You'll need third party translation software to use a glossary - Translation Aggregator is an example of software that allows their use. Your biggest problem will be filling that glossary. I suppose the most straightforward way to do it would be to comb through your MTL results for obvious mistakes and use the glossary to fix them. Do that often enough, and you'll make some progress. There may also be common term glossaries floating around here, if you want to search. E.g. for common fantasy or xuanhuan terms. One of those could give you a kickstart.

    Since I mentioned LNMTL, I should say that's also an option: You can crowdsource glossaries. The easiest way to do that would be to try and have a novel that interests you added to LNMTL, and rely on its existing foundation for crowdsourcing. Otherwise, you can try and start something yourself.
     
  10. UDI

    UDI 『SCP-9369-EggMan』/ Lime Blood / Maid Of Space

    Joined:
    Jul 29, 2017
    Messages:
    305
    Likes Received:
    273
    Reading List:
    Link
    thanks a lot mate that really helps!
     
  11. AliceShiki

    AliceShiki 『Ms. Tree』『Ophi-kun's Survival Teacher』

    Joined:
    Apr 27, 2016
    Messages:
    17,056
    Likes Received:
    70,239
    Reading List:
    Link
    Never tried this one! >.<