ticcky / gibbs_phrase_trees Goto Github PK

Common subtree extraction from analytical parses of sentences using Gibbs sampling.

How to use?
-----------
  # python gibbs_trees_flat.py --treebank sample_input.txt --out sample_output.txt

TSV format
----------
<word id>TAB<word lemma>TAB<word parent id>
...

 - word parent id is id of the parent word or 0 if word's parent is root

Recommendation
--------------
Use Cython or PyPy to process larger treebanks.