tailieunhanh - Báo cáo khoa học: "FROM STRUCTURE-DRIVEN SEPARATE SEMANTIC GENERATION REPRESENTATIONS"

A new approach to structure-driven generation is I)resented that is based on a separate semantics as input structure. For the first time, a GPSGbased formalism is complemented with a system of pattern-action rules that relate the parts of a semantics to appropriate syntactic rules. This way a front end generator can be adapted to some application system (such as a machine translation system) more easily than would be possible with many previous generators based on modern grammar formalisms. | STRUCTURE-DRIVEN GENERATION FROM SEPARATE SEMANTIC REPRESENTATIONS Stephan Busemann Deutsches Forschungszentrum fiir Kiinstliche Intelligenz DFKI GmbH Stuhlsatzenhausweg 3 D-6600 Saarbriicken 11 uucp buscmann@ ABSTRACT A new approach to structure-driven generation is presented that is based on a separate semantics as input structure. For the first time a GPSG-based formalism is complemented with a system of pattern-action rules that relate the parts of a semantics to appropriate syntactic rules. This way a front end generator can be adapted to some application system such as a machine translation system more easily than would be possible with many previous generators based on modern grammar formalisms. 1 INTRODUCTION In the field of unification-based computational linguistics current research on tactical natural language NL generation concentrates on the following problem Given a semantic representation which is often called logical form LF and a grammar that includes a lexicon what arc the surface strings corresponding to the semantic representation A variety of approaches to solving this problem in an efficient way has been put forward on the basis of unification-based grammar formalisms with a context-free backbone and complex categories for some discussion see . Shieber et al. 1990 . Most of this work shares a Montagovian view of semantics by assuming that LF be integrated into the grammar rules thus assigning to each syntactic category its semantic representation. Within this integrated-semantics approach the generation task mainly consists of reconstructing a 1 This work was partially funded by the German Minister for Research and Technology BMFT wider contract ITW 9002. Most of the research underlying this article was accomplished within the EUROTRA-D accompanying research project KIT-FAST at the Technical University of Berlin and funded by the BMFT wider contract 1013211. I wish to thank Christa Hauenschild John Nerbonne and Hans Uszkorcit