This software converts Combinatory Categorial Grammar (CCG) derivations to Phrase Structure Trees (PST). For a full description of the method, and discussion of results, see:

Robust Conversion of CCG Derivations to Phrase Structure Trees, Jonathan K. Kummerfeld, James R. Curran and Dan Klein, ACL (short) 2012

To use the system, download it one of these ways, and run as shown below:

If you use my code in your own work, please cite the paper:

  author    = {Jonathan K. Kummerfeld  and  Dan Klein  and  James R. Curran},
  title     = {Robust Conversion of {CCG} Derivations to Phrase Structure Trees},
  booktitle = {Proceedings of the 50th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)},
  month     = {July},
  year      = {2012},
  address   = {Jeju Island, Korea},
  pages     = {105--109},
  software  = {},
  url       = {},

Running the code

On a sample of CCGbank:

./src/ sample.gold_ptb sample.ccgbank -print_comparison -prefix=sample.ccgbank -verbose -method=markedup ./src/markedup

On a sample of C&C Parser output:

./src/ sample.gold_ptb sample.candc -print_comparison -prefix=sample.candc -verbose -method=markedup ./src/markedup

Conversion output will be in:

The code also comes with a sample of parses from the Penn Treebank section 00, the corresponding parses from CCGbank section 00, and the C&C parser output on the same sentences.

The src directory contains all of the python files required to execute and the instruction sets for categories (in markedup).