How can I make pseudo-projective parsing work? #5519
-
Hi, not sure if this is a bug or not but most likely I am doing something wrong but not sure where to look for answers. As you can see both numbers are dependent on BBBB which makes me think that arcs can't cross and there is no support for non-projective structures (As I said, most likely I am wrong on this)
Thank you so much in advance for any help! My code is:
Your Environment
|
Beta Was this translation helpful? Give feedback.
Replies: 1 comment
-
I'm not sure how large the StackOverflow audience is for pseudo-projective dependency parsing, but let's initially answer this there since it's more of a usage question and people are more likely to find the answer there in the future: https://stackoverflow.com/a/62160956/461847 SO answer copied below: The problem is that the simple training example script isn't projectivitizing the training instances when initializing and training the model. The parsing algorithm itself can only handle projective parses, but if the parser component finds projectivized labels in its output, they're deprojectivitzed in a postprocessing step. You don't need to modify any parser settings (so starting with a German model makes no difference), just provide projectivized input in the right format. The initial projectivization is handled automatically by the train CLI, which uses However, a toy example also works fine as long as you create projectivized training examples (with
|
Beta Was this translation helpful? Give feedback.
I'm not sure how large the StackOverflow audience is for pseudo-projective dependency parsing, but let's initially answer this there since it's more of a usage question and people are more likely to find the answer there in the future: https://stackoverflow.com/a/62160956/461847
SO answer copied below:
The problem is that the simple training example script isn't projectivitizing the training instances when initializing and training the model. The parsing algorithm itself can only handle projective parses, but if the parser component finds projectivized labels in its output, they're deprojectivitzed in a postprocessing step. You don't need to modify any parser settings (so starting with a …