Enriching TimeBank: Towards a more precise annotation of temporal relations in a text

Enriching TimeBank: Towards a More Precise Annotation of Temporal Relations in a Text Volker Gast∗, Lennart Bierkandt∗, Stephan Druskat†, Christoph Rzymski∗ ∗Friedrich Schiller University Jena, English Department, Ernst-Abbe-Platz 8, D-07743 Jena †Humboldt University of Berlin, Department of German Studies and Linguistics, Dorotheenstr. 24, D-10117 Berlin volker.gast@uni-jena.de, post@lennartbierkandt.de, stephan.druskat@uni-jena.de, christoph.rzymski@uni-jena.de Abstract We propose a way of enriching the TimeML annotations of TimeBank by adding information about the Topic Time in terms of Klein (1994). The annotations are partly automatic, partly inferential and partly manual. The corpus was converted into the native format of the annotation software GraphAnno and POS-tagged using the Stanford bidirectional dependency network tagger. On top of each finite verb, a FIN-node with tense information was created, and on top of any FIN-node, a TOPICTIME-node, in accordance with Klein’s (1994) treatment of finiteness as the linguistic correlate of the Topic Time. Each TOPICTIME-node is linked to a MAKEINSTANCE-node representing an (instantiated) event in TimeML (Pustejovsky et al., 2005), the markup language used for the annotation of TimeBank. For such links we introduce a new category, ELINK. ELINKs capture the relationship between the Topic Time (TT) and the Time of Situation (TSit) and have an aspectual interpretation in Klein’s (1994) theory. In addition to these automatic and inferential annotations, some TLINKs were added manually. Using an example from the corpus, we show that the inclusion of the Topic Time in the annotations allows for a richer representation of the temporal structure than does TimeML. A way of representing this structure in a diagrammatic form similar to the T-Box format (Verhagen, 2007) is proposed. Keywords: inferential annotation, Topic Time, GraphAnno 1. Introduction TimeML1 is a markup language which allows us, among other things, to annotate and analyse the temporal structure of a text and to represent it diagrammatically, e.g. in the T-Box format (Verhagen, 2007). It has been used to anno- tate TimeBank (Pustejovsky et al., 2003),2 a collection of news articles from various sources, including the Wall Street Journal. While TimeML is an unrivaled standard in corpus markup, the annotations are somewhat sparse from a theo- retical point of view, as they only provide information about the temporal order of events, but not about the ‘Reference Time’ in terms of Reichenbach (1947), or the ‘Topic Time’ in terms of Klein (1994) (Derczynski and Gaizauskas, 2013; Gast et al., 2015b). In this paper we propose a way of en- riching TimeBank in such a way that it allows us to represent and extract temporal relations in a more precise way. We start in Section 2. with some remarks on tense and as- pect. In Section 3., we describe the annotation procedure, using a sentence from TimeBank for illustration. Section 4. contains the conclusions. 2. Theories of tense and aspect Traditionally, tense is regarded as a relation between some event and the moment of utterance (Leech, 1971). Accord- ingly, the past tense is assumed to be used to describe events that are located before the moment of utterance. However, since Reichenbach (1947) …

Content truncated to 500 words