Renate Henschel

Pronominalization revisited

Pronominalization has been related to the idea of a local focus - a set of discourse entities in the speakers centre of attention, for example in Gundel et al. (1993)s givenness hierarchy or in centering theory. Both accounts say that the determination of the focus depends on syntactic as well as pragmatic factors, but have not been able to pin those factors down. In this paper, we uncover the major factors which determine the focus set in descriptive texts. This new focus definition has been evaluated with respect to two corpora: museum exhibit labels, and newspaper articles. It provides an operationalizable basis for pronoun production, and has been implemented as the reusable module gnome-np. The algorithm behind gnome-np is compared with the most recent pronoun generation algorithm of McCoy and Strube (1999).

Corpus-based NP modifier generation

This paper describes how we annotated and analysed the NP modifiers in a corpus of museum descriptions to discover rules for the selection and realisation of such modifiers, in particular non-referring ones. We implemented the regularities into an extension of the ILEX system to generate complex NPs capable of serving multiple communicative goals.

The Merged Upper Model: a linguistic ontology for German and English

A detailed comparison of the Penman Upper Model and the KOMET German Upper Model has been carried out in order to construct a new Merged Upper Model capable of serving as the ideational basis for generation in both English and German. Previously proposed criteria for conducting such a merge are expanded on and evaluated. It is established that no (semi-)automatic merging of such knowledge sources can be expected to produce a reasonable result and that detailed comparison of the kind reported is essential. The result of the merge is now being used as the basis for sentence generation in English, German and Dutch.

A brief introduction to the GeM annotation schema for complex document layout

In this paper we sketch the design, motivation and use of the GeM annotation scheme: an XML-based annotation framework for preparing corpora involving documents with complex layout of text, graphics, diagrams, layout and other navigational elements. We set out the basic organizational layers, contrast the technical approach with some other schemes for complex markup in the XML tradition, and indicate some of the applications we are pursuing.