Title Automatic Methods to Supplement Broad-Coverage Subcategorization Lexicons
Author(s) Michael Schiehlen, Kristina Spranger

Institute for Computational Linguistics, University of Stuttgart

Session O2-W
Abstract The paper describes a system for extracting subcategorization frames of verbs not found in existing broad-coverage valency lexicons. The system uses two parameters: the results of a finite-state parser and the predictions of a set of automatically learned rules which transfer subcategorization frames from cognate verbs. An in-depth evaluation quantified the contribution of the individual parameters.
Keyword(s) subcategorization extraction, partial parsing, prefix verbs, evaluation, argument--adjunct distinction, supplementing lexicons
Language(s) German
Full Paper 620.pdf