Selva: A New Syntactic Parser for Portuguese

摘要

We describe an ongoing project whose aim is to build a parser for Brazilian Portuguese, Selva, which can be used as a basis for subsequent research in natural language processing, such as automatic translation and ellipsis and anaphora resolution. The parser is meant to handle arbitrary prose texts in unrestricted domains, including the full range of coordination and subordination constructs. The parser operates separately on each sentence, and outputs all its syntactically valid derivation trees.