Using the Tools
The tools can be used in various configurations, either as stand-alone components or in a single pipeline which outputs all the grammatical and semantic information from each tool. The tools expect tokenised input in the style of CCGbank (based on the Penn Treebank), but there is some flexibility in terms of the input the tools can handle.
- Input/Output Formats
- Configurations and Command Line Arguments
- C&C parser
- C&C question parser
- C&C pipeline (including SOAP server)
- Boxer (semantic representations)
- C&C taggers
- Training -- training the CCG parser from CCGbank
- The markedup file -- grammatical relations output
