Susanne Kronenberg

Grammars and Discourse Theory to Describe and Recognize Mechanical Assemblies

In recent years numerous approaches to automatic reasoning about mechanical assemblies have been published. CAD techniques, graph-based approaches and semantic methods to model assembled objects were proposed. However, all these methods are difficult to employ for visually supervising assembly processes since they lack the flexibility, generality or ease required in practice. This paper presents the idea of using syntactic methods known from discourse theory in order to model classes of mechanical assemblies. Moreover, a criterion to derive recursive grammatical production rules is introduced so that the representation of a class of assemblies becomes especially compact. This general representation scheme allows to automatically derive hierarchical structural descriptions of individual assemblies by means of computer vision.

Disambiguation of Utterances by Visual Context Information

Extrapositions are discontinuous constructions of spontaneous speech where constituents are extraposed into the Nachfeld. Parallelism constraints between the source sentence and the extraposed constituents are claimed to govern all possible interpretations induced by an extra-posed constituent. The aim of this article is to integrate visual context information into the parsing model to avoid possible overgeneralizations which might arise from contradictory information given by the source sentence and the extraposed constituent. Therefore, a careful approach to uncertainty is needed because the system is concerned with erroneous, vague, and incomplete data from vision and speech recognition. Bayesian Networks provide an adequate environment to model these uncertain¬ties. The conditional probabilities of the Network are estimated from psycholinguistic experiments.

Inkrementelle Verarbeitung von diskontinuierlichen Konstruktionen in gesprochener Sprache

Der vorgestellte Ansatz zur Verarbeitung von gesprochener Sprache ermoglicht, diskontinuierliche Konstruktionen inkrementell zu parsen. Charakteristisch fur aufgabenorientierte Dialoge sind Konstruktionen, die von Produzent und Rezipient koordiniert produziert werden. Exemplarisch fur solche Konstruktionen werden Extraversionen nach rechts untersucht. Bei diesen werden Konstituenten uber Konstituentengrenzen oder uber die Satzklammer nach rechts extravertiert. Der vorgestellte Ansatz, basierend auf einem LR(1)-Parser und einer Unifikationsgrammatik, bietet eine Moglichkeit, solche Konstruktionen inkrementell zu verarbeiten, indem der referentielle Bezug zwischen extravertierter Konstituente und Bezugskonstituenten nachtraglich erstellt wird.