FSPar – a Cascaded Finite-State Parser for German

FSPar is a rule-based dependency parser. It is based on the approach by Abney (1996) of partial parsing by finite state cascades. FSPar also includes its own internal pipeline of lexically informed tokenizing and lemmatizing and applies the TreeTagger for part-of-speech tagging. It makes use of a large integrated lexical knowledge base and generates underspecified graphs with respect to head selection as well as dependency labels. Additional annotations include: sentence border, token, lemma, part-of-speech tag (Stuttgart-Tübingen-TagSet), morphological information, and some candidates for coreference resolution.

Michael Schiehlen (2003). A Cascaded Finite-State Parser for German. Proceedings of the Research Note Sessions of the 10th Conference of the European Chapter of the Association for Computational Linguistics (EACL'03), 163-166. Budapest, Hungary.