Context-Free Grammars and Constituency Parsing

19.3 Treebanks

1. What is a treebank, and how are treebanks usually built?
   A corpus in which every sentence is annotated with a parse tree. Usually a parser is run over each sentence and linguists hand-correct the output (e.g. the Penn Treebank for English, Arabic, Chinese).
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 19, §19.3, pp. 430–431
2. How does a treebank implicitly define a grammar, and what is notable about the Penn Treebank’s?
   The CFG rules can be read off the parsed sentences. The Penn Treebank grammar is very flat, giving thousands of rules, e.g. separate VP rules for every number and arrangement of PPs and arguments.
   • Jurafsky & Martin, Speech and Language Processing, 3rd ed. draft, ch. 19, §19.3, pp. 431–432