Nicola Mira 

Language Weaver doesn't have access to your translations, or to your translation memories.  Aside from the privacy and security reasons which are important for us to secure, how do we know they are good enough to be used to train the engines?  Using random translations by unknown sources isn't a good idea!

There's a better reply to this question here:

 RE: Confidentiality question regarding Language Weaver 

Nicola Mira 

Language Weaver doesn't have access to your translations, or to your translation memories.  Aside from the privacy and security reasons which are important for us to secure, how do we know they are good enough to be used to train the engines?  Using random translations by unknown sources isn't a good idea!

There's a better reply to this question here:

 RE: Confidentiality question regarding Language Weaver 

Hi Paul,

Thank you very much for the clarification (and for directing me to A. Simon's post), which prompts me to ask another question - with apologies for being perhaps ignorant/naïve: what kind of sources does Language Weaver use to build its TMs? And how does it update and improve on them?

With many thanks again and all best,

N

 Nicola Mira 

Maybe Arnaud Simon is best placed to answer this one.  There will be two basic concepts - trained engines and generic engines.  The trained you can enhance and perfect yourself, the generic will be using our own proprietary technology.  But I'm sure Arnaud can provide a better explanation.

Thank you Paul!