This paper presents an automated empirical evaluation of the relationship between clausal structure and pronominal reference. Past work

The division of text into visual segments such as sentences, paragraphs and sections achieves many

This paper presents DiZer, an automatic DIscourse analyZER for Brazilian Portuguese. Given a source text, the system automatically produces its corresponding rhetorical analysis, following Rhetorical Structure Theory -- RST [1]. A rhetorical repository, which is DiZer main component, makes the automatic analysis possible. This repository, produced by means of a corpus analysis, includes discourse analysis patterns that focus on knowledge about discourse markers, indicative phrases and words usages. When applicable, potential rhetorical relations are indicated. A preliminary evaluation of the system is also presented.

This paper presents an empirical evaluation of a pronoun resolution algorithm augmented with discourse segmentation information. Past work has shown that segmenting discourse can aid in pronoun resolution by making potentially erroneous candidates inaccessible to a pronoun's search. However, implementing this in practice has been difficult given the complexities associated with deciding on a useful scheme and then generating the segmentation reliably. In our study, we investigate whether or not shallow schemes that are easy to generate will improve pronoun resolution in a task-oriented corpus of dialogues. Our results show that incorporating this shallow segmentation at best marginally improves pronoun resolution performance.

Temporal information is essential in many Natural Language Processing applications. The paper presents our activities and research towards obtaining a parallel version, English-Romanian, of the TimeBank 1.2. corpus, annotated according to the TimeML 1.2.1. standard. The automatic import from English to Romanian can be done automatically for 96.53 % of the temporal markups; it will determine the applicability of the general temporal theories when used in Romanian. The parallel corpus will be at the basis of other temporal processing tools.

The paper describes the construction and usage of the Romanian version of the TimeBank corpus. The success rate of 96.53 % for the automatic import of the temporal annotation from English to Romanian shows that the automatic transfer is a worth doing enterprise if temporality is to be studied in another language than the one for which TimeML, the annotation standard used, was developed. A preliminary study identifies the main situations that occurred during the automatic transfer, as well as temporal elements not (yet) marked in the English corpus. 1.

Temporal relations are binary relations that connect temporal events or temporal expressions in texts. On the other hand, texts display discourse structures, as those expressed by Rhetorical Structure Theory (Mann and Thompson, 1988). Different authors suggested that temporality should manifest an intimate correlation with the discourse structure (Asher and Lascarides, 2003). This paper indicates a method that leads to a quantitative investigation of this connection and shows to what extend the link could be used for two purposes: reducing the human effort in manual annotation of temporal relations and improving the automatic annotation of temporal relations. The experimental data we are using for our study consists of twelve texts from the RST Discourse Treebank (Carlson et al., 2003) which have been found to be also annotated for temporal expressions, events and relations, in the TimeBank corpus 1. The original RST annotation evidences RST relations between elementary discourse units (edus) which have the size of a clause or smaller. The temporal annotation marks temporal expressions, events, signals and their links, according to the TimeML standard 2. To each edu in the original RST annotation we add the computed vein expressions