I'm working with a partially structured grammar (for parsing imperative sentences). One problem is that I may say "can you please tell joe that ()" where () can be any english sentence. Is it possible recognize the first part with a grammar and then once it hits some terminal to change into an n-gram model? Or are there better ways to do this (I still want to be able to recognize the second part)? Thanks!
PS: I need to use the grammar so that it is easy for me to extract the action to use.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
This particular problem has been an interest of mine, but Sphinx-4 currently does not provide an implementation to support it. The more modern approach to this problem, however, appears to be to use n-Gram language models coupled with statistical language learning approaches.
Will
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm working with a partially structured grammar (for parsing imperative sentences). One problem is that I may say "can you please tell joe that ()" where () can be any english sentence. Is it possible recognize the first part with a grammar and then once it hits some terminal to change into an n-gram model? Or are there better ways to do this (I still want to be able to recognize the second part)? Thanks!
PS: I need to use the grammar so that it is easy for me to extract the action to use.
Hi:
This particular problem has been an interest of mine, but Sphinx-4 currently does not provide an implementation to support it. The more modern approach to this problem, however, appears to be to use n-Gram language models coupled with statistical language learning approaches.
Will