Information structure in the Potsdam Commentary Corpus: Topics

16215

Authors

nan

Abstract

AbstractThe Potsdam Commentary Corpus is a collection of 175 German newspaper commentaries annotated on a variety of different layers. This paper introduces a new layer that covers the linguistic notion of information-structural topic (not to be confused with ‘topic’ as applied to documents in information retrieval). To our knowledge, this is the first larger topic-annotated resource for German (and one of the first for any language). We describe the annotation guidelines and the annotation process, and the results of an inter-annotator agreement study, which compare favourably to the related work. The annotated corpus is freely available for research.

Keywords

nan

Full Content

Information structure in the Potsdam Commentary Corpus: Topics Manfred Stede and Sara Mamprin Applied Computational Linguistics / FSP Cognitive Science Universit¨at Potsdam (Germany) stede@uni-potsdam.de, mamprin@uni-potsdam.de Abstract The Potsdam Commentary Corpus is a collection of 175 German newspaper commentaries annotated on a variety of different layers. This paper introduces a new layer that covers the linguistic notion of information-structural topic (not to be confused with ‘topic’ as applied to documents in information retrieval). To our knowledge, this is the first larger topic-annotated resource for German (and one of the first for any language). We describe the annotation guidelines and the annotation process, and the results of an inter-annotator agreement study, which compare favourably to the related work. The annotated corpus is freely available for research. Keywords: Information Structure, Annotation, Multi-layer text corpus 1. Introduction The linguistic notion of information structure (henceforth: IS) was conceived to explain certain phenomena of relative prominence in linguistic utterances, both in prosody (into- nation, phrasing) and in syntax (word order, embedding). It also can have effects on truth-conditional semantics. While a number of competing theories and terminologies persist in the literature, there seems to be a minimal consensus that there are three central notions to be distinguished. Be- low we illustrate them with examples taken from (Krifka and Musan, 2012), who provide a thorough overview of the matter. (Upper case letters denote stress on the syllable.) • The information status of discourse referents (often given versus new; sometimes using much more de- tailed scales) (1) Ten years after John inherited an old farm, he SOLD [the shed]given To mark the givenness in speech, the shed needs to be deaccentuated. (Otherwise, the shed would be under- stood as just part of the farm, which renders is non- given.) • The focus/background partitioning of utterances to re- flect prominence. This is typically illustrated with a question representing prior context. (2) A: What did John show Mary? B: John showed Mary [the PICtures]F . A. Who did John show the pictures? B: John showed [MAry]F the pictures. • The topic as the entitiy that an utterance or a sentence “is about”, often in a dichotomy with comment, i.e., the state of affairs that is being predicated of the topic. (3) (a) [Aristotle Onassis]topic married Jacqueline Kennedy. (b) [Jacqueline Kennedy]topic married Aristotle Onassis. Another way of framing these issues is the notion of “in- formation packaging” (Chafe, 1976), which emphasizes the cognitive motivations for making the variety of choices that a speaker faces when formulating linguistic units – it should be done in such a way that the utterance is easy to pro- cess by the addressee, who seeks to mentally reconstruct the message conveyed by the speaker, and its underlying intention. Arguably, at least some aspects of information structure are most fruitfully applied to spoken language, since in many languages prosody is heavily influenced by it, and certain morphosyntactic markers of information-structural aspects occur more often in spoken than in written lan- guage. Nonetheless, especially information status and topic …

Content truncated to 500 words

Labeled Text

AbstractThe Potsdam Commentary Corpus is a collection of 175 German newspaper commentaries annotated on a variety of different layers. This paper introduces a new layer that covers the linguistic notion of information-structural topic (not to be confused with ‘topic’ as applied to documents in information retrieval). To our knowledge, this is the first larger topic-annotated resource for German (and one of the first for any language). We describe the annotation guidelines and the annotation process, and the results of an inter-annotator agreement study, which compare favourably to the related work. The annotated corpus is freely available for research.

Publication Details

Paper ID:
16215
Year:
2016
Publication:
TIPSTER
Publication Type:
conf
DOI:
nan