Evaluation of Broad-Coverage Natural-Language Parsers
- Beyond PARSEVAL
- Satoshi Sekine
- John A. Carroll
- incr tsdb()
The taks of parser evaluation is the measure the speed and accuracy of (symbolic or stochastic) parsers with respect to a manually parsed test corpus. Test data can either be a treebank constructed naturally occurring data, such as the "Wall Street Corpus", or a "test suite" containing constructed examples which cover all interesting linguistic phenomena. An important issue in parser evaluation is the comparison of parsers and grammars based on different linguistic theories. Evaluations have focussed on phrase structure or dependency structures.