How would you get the associated probability for the parse tree returned by the Treebank parser, or get it to return multiple trees?
It seems to create a parsing for any input, even random noise. This is admirable, but I'd like to be able to detect an uncertain parsing (e.g. in the case of random noise) so I can prompt the user for clarification.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Nevermind. Just figured out how to query the parser's arguments, and found the "k" parameter, which returns multiple parses along with their log-probabilities. Exactly what I need!
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
How would you get the associated probability for the parse tree returned by the Treebank parser, or get it to return multiple trees?
It seems to create a parsing for any input, even random noise. This is admirable, but I'd like to be able to detect an uncertain parsing (e.g. in the case of random noise) so I can prompt the user for clarification.
Nevermind. Just figured out how to query the parser's arguments, and found the "k" parameter, which returns multiple parses along with their log-probabilities. Exactly what I need!
What base do the log-probabilities use? I'm having trouble converting them in to a probability normalized between .
e, what Math.log gives you back in java. Hope this helps…Tom