EasyCCG is a Combinatory Categorial Grammar (CCG) parser that implements all combinators and supports parsing to logical form. It provides a common grammar format and includes parameter estimation capabilities for probabilistic CCG models, making it suitable for natural language processing research and applications. ## Key Features - **All Combinators** — Implements the full set of CCG combinators for comprehensive grammatical coverage - **Logical Form Parsing** — Converts parsed sentences into logical representations for semantic analysis - **Probabilistic Parameter Estimation** — Supports training of probabilistic CCG models from data - **Common Grammar Format** — Uses a standardized grammar format for interoperability - **N-best Parsing** — Generates multiple parse candidates with ranking - **Boxer Integration** — Produces Prolog output compatible with the Boxer semantic analysis system - **POS/NER Tag Support** — Accepts pre-tagged input with part-of-speech and named entity recognition tags ## Philosophy EasyCCG is designed as a practical, efficient implementation of CCG parsing that balances theoretical completeness with computational performance, enabling both research experimentation and applied natural language processing tasks.

Stars62

Forks20

Last commit8 years ago