Christa Hauenschild

A constructive view of GPSG or how to make it work

Using the formalism of generalized phrase structure grammar (GPSG) in an NL system (e.g. for machine translation (MT) is promising since the modular structure of the formalism is very well suited to meet some particular needs of MT. However, it seems impossible to implement GPSG in its 1985 version straightforwardly. This would involve a vast overgeneration of structures as well as processes to filter out everything but the admissible tree(s). We therefore argue for a constructive version of GPSG where information is gathered in subsequent steps to produce syntactic structures. As a result, we consider it necessary to incorporate procedural aspects into the formalism in order to use it as a linguistic basis for NL parsing and generation. The paper discusses the major implications of such a modified view of GPSG.

GPSG and German Word Order

In this paper, my main concern is raising questions rather than giving answers. My starting point is Hans Uszkoreit’s revised version of the LP (linear precedence) component (cf. Uszkoreit 1984 and 1986) within the GPSG formalism (cf. Gazdar et al. 1985). I shall discuss some problems of Uszkoreit’s approach that result from the fact that the whole complex phenomenon of German word order is described at a unique level of linguistic representation. I shall then propose a somewhat speculative solution to some of these problems, which is based on a multi-level approach to analysis and generation within the context of machine translation (which is the setting of the project KIT/NASEV and its successor KIT/FAST1).

KI-Methoden in der maschinellen Übersetzung

Die Frage “KI-Methoden in der maschinellen Ubersetzung?” [1] ist mindestens in einer Hinsicht mehrdeutig, namlich bezuglich dessen, was der “Fokus” der Frage (also das eigenlich Erfragte) sein soll. Um nicht den Eindruck einer unkooperativen oder ungeschickten Sender(innen)-Strategie aufkommen zu lassen, will ich in diesem Einleitungsabschnitt die von mir intendierten Interpretationen erlautern, um diese dann zur Strukturierung des gesamten Beitrages zu benutzen. Angesichts der Tatsache, das die Adressaten dieser Ausfuhrungen Mitglieder der KI-Gemeinde sind, ist es naheliegend, “KI-Methoden” (oder zumindest “KI”) als bekannt vorauszusetzen und also als potentiell thematisch anzusehen. Die Interpretation der Titelfrage, die die hochste Praferenz fur den besagten Adressatenkreis haben durfte, ware demzufolge etwa so zu paraphrasieren: “Was macht das Anwendungsgebiet der maschinellen Ubersetzung (Mt fur die XI interessant?” Diese Lesart der Frage wird das Thema von Abschnitt 1 bilden.

A pragmatic concept of theme and rheme for machine translation

The c o n c e p t o f t heme and rheme we w a n t t o p r o p o s e i n t h i s p a p e r i s p r a g m a t i c i n two s e n s e s o f t h e wordz t h e t h e m e / t h e m e s t r u c t u r e (TRS) i s v i e w e d a s a s t r a t e g y o f c o m m u n i c a t i o n a n d t h u s b e l o n g s t o p r a ~ n a t i c s , a n d i t seems t o o f f e r a p r a g m a t i c s o l u t i o n f o r a p r o b l e m o f m a c h i n e t r a n s l a t i o n .